State of the climate in 2013.
TABLE OF CONTENTS List of authors and affiliations Abstract 1. INTRODUCTION 2. GLOBAL CLIMATE a. Overview b. Temperature 1. Surface temperature 2. Lower tropospheric temperature Sidebar 2.1: Temperature extreme indices in 2013. 3. Lower stratospheric temperature. c. Cryosphere 1. Permafrost thermal state 2. Northern Hemisphere continental snow cover extent. 3. Alpine glaciers. d. Hydrological cycle 1. Surface humidity. 2. Total column water vapor. 3. Precipitation. 4. Cloudiness 5. River discharge 6. Groundwater and terrestrial water storage. 7. Soil moisture e. Atmospheric circulation. 1. Mean sea level pressure 2. Surface winds f. Earth radiation budget. 1. Earth radiation budget at top-of-atmosphere 2. Mauna Loa clear-sky atmospheric solar transmission g. Atmospheric composition 1. Long-lived greenhouse gases. 2. Ozone-depleting gases 3. Aerosols 4. Stratospheric ozone 5. Stratospheric water vapor. 6. Tropospheric ozone 7. Carbon monoxide h. Land surface properties. 1. Forest biomass 2. Land surface albedo dynamics 3. Terrestrial vegetation dynamics. 4. Biomass burning 3. GLOBAL OCEANS. a. Overview b. Sea surface temperatures c. Ocean heat content. d. Ocean surface heat and momentum fluxes. e. Sea surface salinity. f. Subsurface salinity g. Surface currents 1. Pacific Ocean 2. Indian Ocean 3. Atlantic Ocean. h. Meridional overturning circulation observations in the North Atlantic Ocean i. Meridional oceanic heat transport in the Atlantic j. Sea level variability and change. k. Global ocean carbon cycle 1. Sea-air carbon dioxide fluxes. 2. Ocean carbon inventory. 3. Anthropogenic ocean acidification. 4. Global ocean phytoplankton 4. THE TROPICS. a. Overview. b. ENSO and the tropical Pacific. 1. Oceanic conditions. 2. Atmospheric circulation c. Tropical intraseasonal activity d. Tropical cyclones. 1. Overview. 2. Atlantic Basin 3. Eastern North Pacific and Central North Pacific Basins 4. Western North Pacific Basin. 5. North Indian Ocean. 6. South Indian Ocean 7. Australian Region Basin 8. Southwest Pacific Basin. e. Tropical cyclone heat potential. f. Global monsoon summary. g. Intertropical convergence zones. 1. Pacific 2. Atlantic. h. Atlantic warm pool. Sidebar 4.1: The 2013 Atlantic hurricane season: blip or flip? i. Indian Ocean dipole. Sidebar 4.2: Super Typhoon Haiyan. 5. THE ARCTIC a. Overview. b. The lower atmosphere: air temperature, clouds and surface radiation. 1. Mean annual surface air temperature 2. Seasonal and regional surface air temperature variability 3. Cloud cover and surface radiation budget Sidebar 5.1: Rapid Arctic warming and midlatitude weather patterns: are they connected? c. Arctic ozone d. UV radiation. e. Carbon dioxide and methane. Sidebar 5.2: Radiative forcing by black carbon in the Arctic. f. Sea ice cover 1. Sea ice extent. 2. Age of the ice. 3. Ice thickness. g. Ocean temperature and salinity 1. Summer sea surface temperature. 2. Upper ocean salinity. 3. Freshwater content Sidebar 5.3: Ocean acidification in the Arctic. 4. Pacific Water layer 5. Atlantic Water layer. h. Terrestrial snow cover 1. Snow cover extent 2. Snow cover duration 3. Snow depth. 4. Snow water equivalent. i. Glaciers and ice caps (outside Greenland) j. Greenland Ice Sheet 1. Satellite observations of surface melting and albedo. 2. Surface mass balance and river discharge. 3. Surface air temperature observations. 4. Satellite observations of ice mass and marine-terminating glaciers. k. Lake ice l. Terrestrial permafrost. 1. Permafrost temperature 2. Active layer thickness. 6. ANTARCTICA. a. Introduction b. Atmospheric circulation c. Surface manned and automatic weather station observations. d. Net precipitation (P - E) e. 2012/13 Seasonal melt extent and duration. f. Sea ice extent, concentration, and duration. g. Ozone depletion Sidebar 6.1: Ultra-low temperatures near dome a, Antarctica 7. REGIONAL CLIMATES a. Overview. b. North America 1. Canada 2. United States Sidebar 7.1: Trends in surface radiation over the United States since the mid-1990s 3. Mexico c. Central America and the Caribbean 1. Central America 2. The Caribbean. d. South America. 1. Northern South America and the tropical Andes 2. Tropical South America east of the Andes 3. Southern South America. Sidebar 7.2: Extreme heat wave over central southern South America during December 2013 . e. Africa 1. Northern Africa. 2. West Africa. 3. Eastern Africa. 4. Southern Africa 5. Western Indian Ocean countries f. Europe and the Middle East. 1. Overview. 2. Central and Western Europe Sidebar 7.3: Intense flooding in central Europe. 3. Nordic and Baltic countries 4. Iberian Peninsula 5. Mediterranean, Italy, and Balkan States 6. Eastern Europe 7. Middle East. g. Asia. 1. Overview 2. Russia. 3. East Asia Sidebar 7.4: Extreme conditions in East Asia in Summer 2013. 4. South Asia 5. Southwest Asia h. Oceania. 1. Overview 2. Northwest Pacific and Micronesia. 3. Southwest Pacific 4. Australia Sidebar 7.5: A year of persistent and widespread heat for Australia. 5. New Zealand. APPENDIX I: Seasonal Summaries APPENDIX 2: Relevant Datasets and Sources ACKNOWLEDGMENTS ACRONYMS AND ABBREVIATIONS REFERENCES
I. INTRODUCTION--D. S. Arndt, J. Blunden, and K. M. Willett
We are pleased to present and be part of this 24th edition of the annual State of the Climate series, which began as NOAA's Climate Assessment, and now completes its 19th consecutive edition associated with the Bulletin of the American Meteorological Society (BAMS).
By fate of the calendar and the synching of multiyear schedules, this is one of several comprehensive reports on the climate system to be released in 2014, following some components of the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, and the United States' National Climate Assessment, and others. This State of the Climate series is unique compared to these larger assessments for its strict focus on the state of the physical climate system, and our capacity to observe it. It consciously defers the attribution of specific events to other studies.
In some ways this effort can be viewed as the "annual physical" of the climate system. Like any annual physical, diagnoses are more robust when based upon data from multiple independent tests, e.g., today's measurements are richer and more meaningful when they are considered in context with past observations; and changes are best detected and monitored when done on a regular basis.
The chapter and section editors are the heart of this series, and have the challenging role of balancing the desire for more and broader content, while constructing the most concise report possible. Global-scale incoming solar radiation is new in this report, as is surface humidity over land areas. Some atmospheric composition sections have been reworked, including the addition of a tropospheric ozone section, following its introduction as a sidebar last year. The Tropics chapter has been refreshed, including new tropical cyclone basin definitions and a global monsoon analysis.
For most of 2013, it appeared that extreme heat and drought in Australia, reminiscent of the United States' experience in 2012, would be the singular and foremost climate story of the year. However, November saw the development, intensification, and catastrophic landfall of Super Typhoon Haiyan. Although Haiyan's meteorological life cycle is measured in weeks, its impact on humanity within and beyond the region will be measured in generations. In a physical climate sense, the seas over which it evolved have undergone changes on the scales of climate variability and change. SSTs were record warm in much the region, and the region's sea level rise has outpaced the global average in recent years. For these reasons, it was most appropriate for the front and back covers of this report to reflect its importance as a meteorological, climatological, and human event.
On the topic of the intersection of climate and humanity, we thank and recognize our authors throughout the world who organize the data and create the content of this report. We are proud of the ever-increasing international participation in this series. Many of our authors, and the institutions they represent, must deal with disruptions in the course of their work. Despite the demise of observing systems, and in the face of budgetary pressures, and even civil unrest, their contributions make this report and our profession more whole. We are fortunate to share this space as we share our climate system itself--with so many fine colleagues.
An overview of findings is presented in the Abstract, Plate 1.1, and Fig. 1.1. Chapter 2 features global-scale climate variables; Chapter 3 highlights the global oceans; and Chapter 4 covers tropical climate phenomena including tropical cyclones. The Arctic and Antarctic respond differently through time and are reported in separate chapters (5 and 6, respectively). Chapter 7 provides a regional perspective authored largely by local government climate specialists. Sidebars included in each chapter are intended to provide background information on a significant climate event from 2013, a developing technology, or emerging dataset germane to the chapter's content. A list of relevant datasets and their sources for all chapters is provided as an Appendix.
ESSENTIAL CLIMATE VARIABLES
Time series of major climate indicators are again presented in this introductory chapter. Many of these indicators are essential climate variables (ECVs), originally defined in GCOS 2003 and updated again by GCOS in 2010.
The following ECVs, included in this edition, are considered "fully monitored", in that they are observed and analyzed across much of the world, with a sufficiently long-term dataset that has peer-reviewed documentation:
* Atmospheric Surface: air temperature, precipitation, air pressure, water vapor, wind speed and direction.
* Atmospheric Upper Air: earth radiation budget, temperature, water vapor.
* Atmospheric Composition: carbon dioxide, methane, other long-lived gases, ozone.
* Ocean Surface: temperature, salinity, sea level, sea ice, current, ocean color, phytoplankton.
* Ocean Subsurface: temperature, salinity.
* Terrestrial: snow cover, albedo.
ECVs in this edition that are considered "partially monitored", meeting some but not all of the above requirements, include:
* Atmospheric Upper Air: cloud properties.
* Atmospheric Composition: aerosols and their precursors.
* Ocean Surface: carbon dioxide, ocean acidity.
* Ocean Subsurface: current, carbon.
* Terrestrial: soil moisture, permafrost, glaciers and ice caps, river discharge, groundwater, ice sheets, fraction of absorbed photosynthetically- active radiation, biomass, fire disturbance.
ECVs that are expected to be added in the future include:
* Atmospheric Surface: surface radiation budget.
* Atmospheric Upper Air: wind speed and direction.
* Ocean Surface: sea state.
* Ocean Subsurface: nutrients, ocean tracers, ocean acidity, oxygen.
* Terrestrial: water use, land cover, lakes, leaf area index, soil carbon.
2. GLOBAL CLIMATE--K M. Willett, A. J. Dolman, D. F. Hurst, J. Rennie, and P. W. Thorne, Eds.
a. Overview--P. W. Thorne, A. J. Dolman, D. F. Hurst, J. Rennie, and K. M. Willett
After several years strongly influenced by either La Nina or El Nino events, 2013 was the first full year without either of these phenomena present. Without the typical large-scale dynamical teleconnections driven by variability within the El Nino-Southern Oscillation (ENSO), 2013 was dominated by patterns of regional extremes of temperature and especially the hydrological cycle. Regional variations were particularly notable in the Northern Hemisphere extratropics where anomalously meridional atmospheric circulation occurred throughout much of the year, leading to marked regionally coherent extremes of heat/cold and dry/wet.
This year, an analysis of temperature extremes since 1950 is introduced in Sidebar 2.1, since societal impacts are more often related to extreme events than changes in the mean climate. The year 2013 ranked within the top 10 years for the frequency of warm days and bottom 10 years for the frequency of cool days. The global average maximum temperature index which tracks extreme daytime heat was also within the top 10 highest years, largely driven by the record-warm summer in Australia.
New sections include an updated solar transmission record with a view to subsequent global analyses. New data products have also been introduced. The JRA-55 reanalysis now extends back to 1958, providing a second long-term reanalysis for comparison. The HadISDH surface humidity product now includes land relative humidity and concurs with previous indications of declining relative humidity over land since 2000. Conversely, it was not possible to update total column water vapor from both ground-based GPS and radiosondes this year.
Globally, 2013 was again one of the 10 warmest years on record, both at the surface and in the troposphere, according to the large range of available estimates; however, there is uncertainty in the precise rankings of any given year.
Long-lived greenhouse gases carbon dioxide (C[O.sub.2]), methane (C[H.sub.4]), and nitrous oxide ([N.sub.2]O) continued to increase in the atmosphere during 2013. On 9 May, for the first time since C[O.sub.2] measurements began in 1958 at Mauna Loa, Hawaii, the daily average mole fraction exceeded 400 parts per million (ppm). The global abundance of tropospheric ozone, a greenhouse gas and pollutant, also continued to grow but the root causes have not yet been identified. Global atmospheric burdens of ozone-depleting chlorofluorocarbons (CFCs) continued to decline while those of their replacements increased. Stratospheric ozone levels remain well below the pre-1980 benchmark of ozone layer "recovery". Stratospheric water vapor abundance declined in 2013 after a 6-7 year period of increase that followed the rapid drop in 2000. Aerosol optical depth and carbon monoxide (CO) column measurements by satellites continue to demonstrate the importance of boreal and tropical biomass fires as global sources.
Early indications from the limited available sample of global glaciers are that 2013 was the 24th consecutive year of net glacier loss globally. Snow cover continued to decline in the Northern Hemisphere, as did soil moisture, albeit less strongly due to the absence of a strong ENSO signal. Noticeable also were declines in the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) over north east Brazil (also a region of a dry soil anomaly) and Siberia. Global fire activity was relatively low during 2013, but regionally strong fire activity caused severe air pollution problems in some regions, including Siberia and Sumatra.
Where available, time series and anomaly maps of the variables described in this chapter are shown in Plates 1.1 and 2.1, respectively. Most anomalies are referenced against a standard 1981-2010 climatology which covers the period of many satellite observations and all reanalysis products. Many sections refer to online figures that can be found here: http://dx.doi.Org/10.1175/2014BAMSStateoftheClimate.2.
I) SURFACE TEMPERATURE--A. Sanchez-Lugo, J. J. Kennedy, and P. Berrisford
The year 2013 was the warmest since 2010 and among the 10 warmest years since records began in the mid- to late 1800s. According to four independent observational analyses (NASA-GISS, Hansen et al. 2010; HadCRUT4, Morice et al. 2012; NOAA-NCDC, Smith et al. 2008; IMA, Ishihara 2006), the 2013 globally-averaged surface temperature was 0.20[degrees]-0.21[degrees]C (Table 2.1) above the 1981-2010 average and ranked from second to sixth warmest, depending on the dataset considered.
Each of the four independent teams analyzes air temperatures from weather stations over land and sea surface temperatures (SST) observed from ships and buoys and then merges the land and SST datasets to form a global analysis. While their methods differ, leading to minor differences in anomalies and ranks, all four analyses are in close agreement (Fig. 2.1). The main difference between the four datasets is how each methodology treats areas with little to no data (see Kennedy et al. 2010 for additional details). Recent research (Cowtan and Way 2014) suggests that the current method used to estimate global temperatures for HadCRUT4 might have led to a slight underestimation of global temperatures in recent years.
Global average temperatures are also estimated using reanalyses, which blend model and observational data together. Surface temperatures from reanalyses are the result of in situ and satellite, surface and upper-air, current and prior observations that are propagated by the model. Although errors arise both from data handling and the model, surface temperatures from reanalyses are consistent with observations in regions of good observational coverage at the surface, due in part to the large volumes of assimilated observations (a total of more than 40 billion to date in ERA-Interim). In ERA-Interim (Dee et al. 2011), the 2013 globally-averaged analyzed 2-m temperature was 0.18[degrees]C above the 1981-2010 average (Table 2.1) and ranked as the fourth warmest year in the record, which began in 1979.
Unlike 2011 and 2012 when La Nina conditions were present, neutral ENSO conditions prevailed throughout 2013. The presence of neutral ENSO conditions during 2013 contributed to a global average temperature slightly higher than the two previous years (Fig. 2.1), for all datasets. Typically, the presence of a La Nina tends to have a cooling influence on global temperatures.
Overall, the observational datasets indicate that the annual 2013 surface temperatures were warmer than average across most of the world's land and ocean surfaces (Plate 2.1c; Online Fig. S2.1), with cooler-than-average conditions across central Canada, the central and eastern parts of United States, the equatorial Pacific Ocean, and along the Pacific coast of South America. The globally-averaged annual temperature over land was 0.32[degrees]-0.38[degrees]C above the 1981-2010 average--between the third and sixth warmest land temperature on record, depending on the dataset. The globally-averaged annual temperature over the ocean was 0.13[degrees]-0.17[degrees]C above average--between the fourth and eighth warmest sea surface temperature on record.
ERA-Interim also shows warmer-than-average conditions over many regions of the world (Fig. 2.2). The globally-averaged analyzed 2-m temperature over land was 0.42[degrees]C above average, making 2013 the fourth warmest year in the 35-year period of record. Over ocean it was 0.08[degrees]C above average--the ninth warmest year.
2) LOWER TROPOSPHERIC TEMPERATURE--J. Christy
The 2013 annual, globally-averaged temperature of the lower troposphere (the bulk atmosphere below 10-km altitude or roughly the lower 70% by mass) was approximately +0.2[degrees]C above the 1981-2010 average. This placed 2013 between the fourth and ninth warmest of the past 35 years depending on dataset and about 0.2[degrees]C cooler than the warmest year, 1998 (Fig. 2.3).
Direct measurement of the lower-atmospheric bulk temperature utilizes radiosonde datasets with available data since 1958 and satellites since late 1978. Radiosondes are balloon-borne instrument packages that monitor the temperature of the air as the balloon ascends. Satellites monitor the intensity of upwelling microwave emissions of atmospheric oxygen that are proportional to temperature. Temperature variations in the troposphere are relatively large-scale when averaged over a month or year due to mixing by atmospheric circulation. Lienee, fewer spatial points (i.e., radiosondes) are needed to create a reasonable estimate of the global average compared to the more inhomogeneous surface. However, as at the surface, the instrumentation on both radiosondes and satellites has changed through the years and adjustments to the data are required to produce homogeneous time series for climate analysis. The different choices made in homogenizing the raw series lead to slightly different results (termed "structural uncertainty" below.)
The global temperature anomaly at any point in time is closely tied to the phase of ENSO. Although 2013 was an ENSO-neutral year, lower tropospheric temperatures show some indication of a warm ENSO phase at the end of 2012, which warmed tropical temperatures in the first part of the year (see Fig. 2.23; section 2el). The rest of the year was neutral, with tropical temperatures staying near the 30-year average. Monthly departures from the global average were most positive in January, June, September, and October and least positive in April and May.
Regionally, warm anomalies extended from northeast Asia through the North Pacific, as well as from North Africa to western Russia. Australia experienced its warmest annual anomaly in the upper air record. Antarctica, along with the South Pacific, was also warmer than average. Smaller areas of cooler-than-average temperatures prevailed in central North America, Hawaiian region, eastern Asia, and scattered areas in the Southern Hemisphere oceans (Plate 2.1b).
The long-term global trend based on radiosondes [starting in 1958 (excluding HadAT2)] is +0.15[degrees]C decade1 and both radiosondes and satellites (starting in 1979) is +0.13 [+ or -] 0.02[degrees]C [decade.sup.-1] (Table 2.2). The range represents the variation among the different datasets and serves as an estimate of structural uncertainty in Fig. 2.3. When taking into account the magnitude of the year-to-year variations, there is a statistical confidence range of [+ or -] 0.06[degrees]C [decade.sup.-1], so the trends are still significantly positive. Major volcanic events in 1963, 1982, and 1991 led to periods of cooler temperatures that affected the early and middle part of the tropospheric record, especially in the satellite era, enhancing the upward trend to some extent. A number of smaller eruptions have occurred since 2000 which may also have moderated the temperature slightly (see section 2f2).
Given basic lapse-rate theory (Christy 2013) tropospheric trends are expected to exceed those at the surface by about 1.2 (1.4) for the globe (tropics). Over the satellite era trends in observations indicate a troposphere-to-surface ratio close to or slightly less than 1.0 for both global and tropical domains (Table 2.2). Over the longer radiosonde era the observed results are nearer theoretically-expected values.
In addition to radiosonde and satellite estimates, four reanalyses products are also shown (Fig. 2.3). There is reasonable agreement in the interannual variability and long-term trend between the reanalyses and observation products. ERA-Interim shows good agreement with satellite estimates and is used here to provide the spatial depictions (Plate 2.1b; Fig. 2.4).
3) LOWER STRATOSPHERIC TEMPERATURE--C. S. Long and J. Christy
The global-average temperature in the lower stratosphere for 2013 was slightly below the 1981-2010 climatology (Fig. 2.5). All the various measurement systems (radiosonde, satellite, reanalysis) determined that 2013 was warmer than 2012 and continued a near-neutral to very gradual warming trend from 1995 to present. The annually-averaged temperature anomaly was positive in most regions poleward of the midlatitudes and in the equatorial zone (see Plate 2.1a; Online Fig. S2.2). The subtropics and lower midlatitudes were cooler than the climatology, with a region of negative temperature anomalies extending from central Canada to Alaska in the Northern Hemisphere, as well as southern South America to New Zealand in the Southern Hemisphere. Between these two features were positive temperature anomalies over the eastern Pacific Ocean. These annual features were results of strong monthly departures. The northern polar region experienced positive temperature anomalies in January and February resulting from a stratospheric warming in early January. The central Canada to Alaska negative temperature anomalies arose from below-average temperatures in this region during November and December. The southern polar positive temperature anomalies resulted from above-average polar temperatures during September-December (see Online Fig. S2.7). These warm temperatures resulted in a smaller-than-average ozone hole for 2013 (sections 2g4,6g). The Southern Hemispheric polar circulation was displaced off the South Pole for much of 2013, resulting in negative temperature anomalies from the southern South America to New Zealand region during August-December. The tropical temperature anomalies transitioned from slightly negative in the early months of the year to positive by the end of the year (see Online Figs. S2.3-S2.7), as a result of the quasi-biennial oscillation winds transitioning from descending easterlies to descending westerlies throughout the year.
Historically, the radiosondes (RAOBCORE, Haimberger et al. 2012; RICH, Haimberger et al. 2012; HadAT2, Thorne et al. 2005; and RATPAC, Free et al. 2005) and later on the satellites (RSS, Mears and Wentz 2009; STAR, Zhou and Wang 2010; and UAH, Christy et al. 2011) show a cooling trend from the 1960s through the mid-1990s (Fig. 2.5). Both measuring systems indicate a change in that cooling trend around 1995. Since then there has been a near-neutral or warming trend, depending upon the data product. In prior reports, a linear trend from 1979 to present was determined, because long period time series are necessary for robust assessment of long-term trends. However, it is clear that using one linear trend to characterize the entire period does not accurately describe the behavior since 1995. Four of the most recent meteorological reanalyses (MERRA, Rienecker et al. 2011; ERA-Interim, Dee et al. 2011; JRA-55, Ebita et al. 2011; NCEP-CFSR, Saha et al. 2010) show general agreement with the radiosonde and satellite temperature anomalies. Table 2.3 shows the trends of the various estimates for 1958-95 (radiosonde only), 1979-95, 1995-2013, and 1979-2013. These periods reflect the cooling trend through 1995 and the neutral-to-warming trend from 1995 to present.
More detail about the tropical lower stratospheric trends is obtained if the zonal trends are examined. Figure 2.6 shows the zonal trends for the cooling 1979-95 period and the neutral-to-warming trend from 1995-present. The equatorial region exhibits the smallest trends (both warming and cooling) while the polar latitudes exhibit the greatest trends. The increasing size of the ozone hole and its associated cooling in September-November is a strong contributing factor in the Southern Hemisphere 1979-95 cooling trend. The cooling in the Northern Hemisphere polar region in 1979-1995 can be attributed, to a large extent, to reducing occurrences of winter stratospheric warmings. From 1995 to present there have been more frequent stratospheric warmings and hence a general warming trend in the Northern Hemisphere polar latitudes (Pawson and Naujokat 1999; Manney et al. 2005; Butler and Polyani 2011). In the Southern Hemisphere, there have recently been years with active wave activity during the winter/spring and consequently, warmer average polar temperatures (and smaller associated ozone holes). Between the two polar regions there has been a very slight warming trend.
1) Permafrost thermal state--J. Noetzli, H. H. Christiansen, M. Gugliemin, V. E. Romanovsky, N. I. Shiklomanov, $. L. Smith, and L. Zhao
The Global Terrestrial Network on Permafrost (GTNP) brings together long-term records on ground temperatures and active layer depths from permafrost regions worldwide in order to document the state and changes of permafrost on a global scale.
Arctic permafrost temperatures generally vary from 0[degrees] to -2.5[degrees]C within the discontinuous zone, with colder conditions in the continuous zone from -3[degrees]C in high Arctic Svalbard (Christiansen et al. 2010; Fig. 2.7) to -15[degrees]C elsewhere in the high Arctic (Romanovsky et al. 2010a). Permafrost has warmed over the past two to three decades, and generally continues to warm across the circumpolar north. Record-high temperatures were observed in 2012-13 in the Alaskan Arctic and the Canadian Archipelago (Romanovsky et al. 2013a,b; a detailed discussion of Arctic permafrost is provided in section 51).
Permafrost in the European Alps is discontinuous or patchy and generally warm with temperatures between 0[degrees] and -3[degrees]C (Haeberli et al. 2010; PERMOS 2013; Fig. 2.7). However, on shaded slopes at high elevations permafrost can be as cold as in the high Arctic in Svalbard. So far the lowest borehole temperatures of -5[degrees]C were measured in horizontal boreholes installed in 2008 in a near-vertical rock pillar in the French Mont Blanc Massif at 3800 m above sea level on the Aiguille du Midi (Magnin et al. 2014, manuscript submitted to The Cryosphere). In addition, measurements on the south and north face confirm large temperature differences of up to 6[degrees]C between steep north and south slopes of midlatitude mountains. This results in 3D temperature patterns and large lateral heat fluxes (Noetzli and Gruber 2009). Decadal records for European mountain permafrost show a warming trend at depths of 20 m and more for many but not all sites, especially in the past five years; with smaller increases where permafrost is close to 0[degrees]C (Isaksen et al. 2007; PERMOS 2013) At temperatures close to the melting point phase change processes absorb a part of the energy transported from the atmosphere to the subsurface. That way latent heat can mask atmospheric warming in the underground. At 10-m depth seasonal variations are well displayed, which reveals warmer winters at warmer sites in recent years. Pronounced warming trends are observed in Scandinavia (Isaksen et al. 2011; Fig. 2.7), which are consistent with changes in air temperatures.
In the warm permafrost of the higher altitudes of central Asia, ground temperatures have increased by up to 0.5[degrees]C decade*1 since the early 1990s. Additional boreholes were recently installed in the Qinghai-Xizang Plateau (Zhao et al. 2011) and Mongolia (Sharkhuu and Sharkhuu 2012) as part of GTN-P. The average warming rate of permafrost in these regions was about 0.31[degrees]C decade1 from 1998 to 2010 (Zhao et al. 2011).
The latitudinal transect in maritime Antarctica ANTPAS (Antarctic Permafrost, Soils and Periglacial Environments) was upgraded in 2012 and 2013 with new boreholes near Palmer Station and at Signy Island. Permafrost temperature at 17-m depth was -1.3[degrees]C and active layer thickness (ALT) reached almost 4 m in 2013. Farther south at Rothera station (67[degrees]S), Guglielmin et al. (2014a) reported permafrost temperatures were around -3[degrees]C in 2013.
Changes in ALT vary by region (Shiklomanov et al. 2012), but it is generally increasing globally (Fig. 2.8). The majority of the time series have a length of 10 years beginning around 2000, a few date back as far as 1990. In 2013, ALT was greater than the 1995-2013 mean and similar to or greater than 2012 in some areas, for example, Alaska, Siberia, and Russian European North (see section 51). In eastern Siberia, however, ALT in 2013 was less than average (see section 51). Increases in ALT since the late 1990s have also been observed on Svalbard and Greenland, but these are not spatially and temporally uniform (Christiansen et al. 2010). Elere and also in the European Alps, ALT was similar or lower in 2013 compared to 2012. In the European Alps ALT over the past five years has been greater than measured previously, with new record values in 2012 or 2013 at some of the sites. A general increase in ALT has also been observed in central Asia (e.g., Zhao et al. 2010). Based on the monitoring results extended by a freezing-thawing index model, the average increase of ALT was about 1.33 cm [year.sup.-1] from 1980 to 2010 (Li et al. 2012). At the new sites in maritime Antarctica ALT varied between 0.76 m and 1.4 m. In Victoria Land, continental Antarctica, observations by Guglielmin et al. (2014b) confirmed the thickening of the active layer since 1997 reported by Guglielmin and Cannone (2012), which mainly results from increasing solar radiation in austral summer.
2) Northern Hemisphere continental snow cover extent--D. Robinson
Annual snow cover extent (SCE) over Northern Hemisphere (NH) lands averaged 25.5 million [km.sup.2] in 2013. This is 0.3 million [km.sup.2] more than the 44year average, and ranks 2013 as having the 13th most extensive cover on record (Table 2.4). This evaluation includes the Greenland ice sheet. SCE in 2013 ranged from 49.2 million [km.sup.2] in January to 2.9 million [km.sup.2] in August. Monthly SCE is calculated at the Rutgers Global Snow Lab from daily SCE maps produced by meteorologists at the National Ice Center (a US joint NOAA, Navy, and Coast Guard facility), who rely primarily on optical satellite imagery to construct the maps.
Monthly mean anomalies varied considerably across the year in all regions (Fig. 2.9). The first four months of 2013 saw above-average snow cover extent over Eurasia (EU) and North America (NA). EU SCE observed its sixth snowiest January of the past 47 years. North American SCE exhibited the largest positive anomalies in March (sixth largest for the month) and April (third largest for the month). As seen on multiple occasions over the past decade, May and June SCE were well below the long-term average. NH SCE plummeted from the ninth most extensive in April to the third least extensive coverage in May and second least in June.
Snow arrived early over the Northern Hemisphere continents during fall 2013. Hemispheric rankings were sixth and seventh most extensive in September and October, respectively. The advance of the seasonal snowpack continued at a rapid pace over North America in November (third most extensive) and December (seventh most expensive). However the SCE advance slowed considerably over Eurasia in November, the rank falling to 31st most extensive, though it became above average (20th most extensive) in December.
Unlike the previous winter, contiguous United States SCE in early 2013 was above average. This included the fifth most extensive SCE for April, prior to a rapid melt that left May with the second least extensive SCE on record. Late in 2013, SCE developed quickly and ranked between 5th and 12th most extensive from October through December.
Maps depicting daily, weekly, and monthly conditions, daily and monthly anomalies, and monthly climatologies for the entire period of record may be viewed at the Rutgers Global Snow Lab website (http://snowcover.org). Monthly SCE for the NH, EU, NA, the contiguous United States, Alaska, and Canada are also posted, along with information on how to access weekly areas and the weekly and monthly gridded products.
3) Alpine glaciers and ice sheets--M. Pelto
The World Glacier Monitoring Service (WGMS) record of mass balance and terminus behavior (WGMS 2013) provides a global index for alpine glaciers. Glacier mass balance is the difference between accumulation and ablation. Mass balance was -638 mm in 2012, negative for the 23rd consecutive year. Preliminary data for 2013 from Austria, Canada, Nepal, New Zealand, Norway, and United States indicate it is highly likely that 2013 will be the 24th consecutive year of negative annual balances.
Alpine' glaciers have been studied as sensitive indicators of climate for more than a century, most commonly focusing on changes in terminus position and mass balance. The worldwide retreat of mountain glaciers is one of the clearest signals of ongoing climate change (Haeberli et al. 2000). The retreat is a reflection of strongly negative mass balances over the last 30 years (WGMS 2013).
The cumulative mass balance loss since 1980 is 14.9 m w.e. (meters in water equivalent), the equivalent of cutting a 16.5 m thick slice off the top of the average glacier (Fig. 2.10). The trend is remarkably consistent from region to region (WGMS 2013). WGMS mass balance results based on 30 reference glaciers with 30 years of record are not appreciably different, -15.1 m w.e. The decadal mean annual mass balance was -198 mm in the 1980s, -382 mm in the 1990s, and -740 mm for 2000s. The declining mass balance trend during a period of retreat indicates alpine glaciers are not approaching equilibrium and retreat will continue to be the dominant terminus response. The recent rapid retreat and prolonged negative balances have led to some glaciers disappearing and others fragmenting (Fig. 2.11; Pelto 2010; Carturan et al. 2013).
In 2013 the Austrian glacier inventory examined 96 glaciers: 93 were in retreat, 1 was advancing, and 2 were stationary, with an average terminus change of -17 m. Mass balance in 2013 was slightly negative on three glaciers with completed data. A 170-m increase in annual equilibrium line altitude on 43 glaciers in the Alps from 1984 to 2010 is driving the ongoing retreat (Rabatel et al. 2013).
In Norway terminus fluctuation data from 33 glaciers for 2013 with ongoing assessment indicate 26 retreating, 4 stable, and 3 advancing, with an average terminus change of -12.5 m (Elverhoi 2013). Mass balance surveys with completed results are available for six glaciers; all have negative mass balances with an average loss exceeding 1 m w.e. (Andreassen 2013). Of the five outlet glaciers examined from 2002-13 all retreated; the mean retreat was 190 m.
In the North Cascades, Washington, the 2013 winter accumulation season featured 93% of average (1984-2013) snowpack. The melt season was exceptional with the average June-September temperature tied as the highest for the 1989-2013 period and also having the highest average minimum daily temperatures. This contributed to significant negative balances on all 10 glaciers observed, with an average of-1 m w.e. (Pelto 2013). In British Columbia, end-of-summer snowlines were higher than normal and annual mass balance was significantly negative. In Alaska, all four glaciers with mass balance assessed had significant negative mass balances (Pelto 2013).
In New Zealand, the annual end-of-summer snowline survey on 50 glaciers found snowlines that were slightly above the elevation for glacier equilibrium. Heavy snow accumulation during October was offset by a warm, dry summer with high ablation (NIWA 2013).
In Nepal, the mass balances of Yala, Mera, and Pokalde Glaciers were near equilibrium. Accumulation was the highest of the last seven years, with particularly heavy snow from extratropical storm Phailin (ICIMOD 2013).
d. Hydrological Cycle
I) SURFACE HUMIDITY--K. Willett, A. Simmons, and D. Berry
Over land, specific humidity (q) for 2013 was just above the 1979-2003 average and slightly higher than in 2012, as shown by the in situ HadlSDH and ERA-40/Interim (spatially matched to HadlSDH) and JRA-55 reanalysis products (Fig. 2.12a,b). Over the ocean, specific humidity in 2013 was considerably above the average and slightly higher than 2012, as shown by the NOCSv2.0 in situ product masked to regions where data quality is sufficient (essentially the Northern Hemisphere; Fig. 2.12c). While agreement between and within both in situ and reanalyses product types is generally good over land, there is much less agreement over ocean (Fig. 2.12c,d), though the reanalyses show some similarity in interannual behavior to each other. The differences between the in situ data products and the reanalyses may be partially due to the significant difference in spatial coverage. In regions of poor data coverage the reanalyses still provide values, however, these are less constrained compared to well-sampled regions and the uncertainties are higher. The HOAPS satellite ocean humidity product shows good agreement with NOCSv2.0 in the early period, deteriorating from 1998 where the ENSO signal is not present in HOAPS.
Overall there was more water vapor than average in the near-surface atmosphere in 2013; the long-term behavior of all estimates suggests an increasing trend with more water vapor in the near-surface atmosphere now than in the 1970s.
Plate 2.1f (and Online Figs. S2.8, S2.9) shows the annual average q anomalies over the globe for 2013. The picture is mixed but coherent regional patterns emerge. The southern and central United States, western Europe, central and east Asia, central South America, southern Africa, and central eastern Australia were drier than the 1981-2010 average. These signals stretch across coastlines, showing good consistency between the independent HadlSDH and NOCSv2.0 products, and are also in agreement with ERA-Interim (Online Fig. S2.9). Overall, there are more regions showing moister-than-average anomalies. The midlatitude northern Pacific and northern Atlantic, Southeast Asia, and most of the observed tropical Africa are areas of considerable moistening. Seasonal variability in these features is large (Online Fig. S2.10), with the dry anomalies predominant in the boreal winter (DJF) and spring (MAM).
Over land, relative humidity (RH) in 2013 was far below average although slightly higher than in 2012, as shown by the in situ only HadlSDH and by the ERA-40/Interim and JRA-55 reanalyses (Fig. 2.12e,f). Over ocean, the only estimates for 2013 available are provided by reanalyses. These suggest no significant deviation from average and little variability in the record overall (Fig. 2.12g,h). There is generally greater variability both between and within the in situ and reanalyses estimates for RH. For the in situ record, this suggests high sensitivity to both methodological choice and station coverage which differs between the Dai, HadCRUH, and HadlSDH datasets.
Overall, this means that although there is more water vapor in the near-surface atmosphere relative to the 1970s, the surface atmosphere over land is less saturated. This feature has really only become apparent since -2000 but is clear in all estimates. The drivers of this are not yet fully understood but differences in the rate of warming between land and ocean have been suggested (Simmons et al. 2010) and land surface water availability may be a factor.
Plate 2.1g (and Online Fig. S2.ll) shows the annual average RH anomalies over the globe for 2013. Below-average saturation is a predominant feature of the midlatitude land masses whereas the higher latitudes and tropics, especially India, are more humid than the long-term average. Anomalies are much larger over land than over the oceans, in agreement with ERA-Interim (Online Fig. S2.ll). Seasonal variability in these features is large (Online Fig. S2.12), although the dry regions of central South America, southern Africa, and central eastern Australia, and humid regions of India persist year round.
The majority of products used here to show estimates of surface humidity are described in Willett et al. (2013). This year JRA-55 is shown, which extends the previous JRA-25/JCDAS back to 1958. JRA-55 improves on the JRA-25/JCDAS post-October 2011 record which previously suffered from a large discontinuity due to the absence of precipitable water retrievals from the microwave imagers used. HadISDH has been updated to HadISDH.188.8.131.523p which now includes land RH in addition to land q and different homogenisation methodology (Willett et al. 2014, manuscript submitted to Climate Past). There is negligible difference between the two versions for large scale averages of q.
2) Total column water vapor--C. Mears, S. Ho, L. Peng, and J. Wang
The map of total column water vapor (TCWV) anomalies for 2013 (Plate 2.1e) was made by combining data both from satellite-borne microwave radiometers over ocean (Wentz 1997; Wentz el al. 2007) and COSMIC GPS-RO over land (Ho et al. 2010a,b; Teng et al. 2013; Huang et al. 2013). Despite the lack of any significant ENSO event during 2013, La Nina-like dry anomalies persisted across the central Pacific. There were also pronounced dry anomalies in eastern Australia and both northern and southern Africa. There were pronounced wet anomalies in the eastern and western tropical Pacific, the South Pacific convergence zone, and the Amazon basin. Many of the same features were present in the 2013 precipitation anomalies (Plate 2.1h). The pattern in TCWV over the ocean is confirmed by COSMIC ocean measurements.
Over the ocean, the TCWV anomaly time series (Fig. 2.13a) from the microwave radiometers shows maxima in 1987-88, 1997-98, and 2009-10, associated with El Nino events, as well as a more subtle increasing trend corresponding with increasing global temperatures. A linear fit to this time series suggests that the total amount of vapor over the oceans has increased -3% since 1988. Minima are apparent in Northern Hemisphere winters during the La Nina events of 1988-89, 1992-93, 1999-2000, 2007-08, and late 2010 to mid-2012. Global water vapor has increased since this last minimum. The ocean-only COSMIC data are in general agreement with the radiometer data, but show less of a peak in 2009-10. An increase since late 2010 is also shown in the COSMIC data over land (Fig. 2.13b). A Hovmoller plot derived from the satellite radiometers (Fig. 2.14) shows that the long-term increase in TCWV is occurring at all latitudes, with less variability outside the tropics.
3) PRECIPITATION--R. S. Vose, K. Hilburn, X. Yin, M. Kruk, and A. Becker
Globally, precipitation over land surfaces was near the 1961-90 average in 2013 (Fig. 2.15a). This conclusion is based primarily on station records in the Global Historical Climatology Network (GHCN) Monthly version 2 (Peterson and Vose 1997), which was about 1 mm above normal, and the Global Precipitation Climatology Centre (GPCC) Monitoring Product version 4 (Becker et al. 2013), which was about 1 mm below normal. Historically, GHCN and GPCC have been similar on an annual basis, though GHCN has higher interannual variability due to its smaller network. Land data for a blended satellite-in situ product, the Global Precipitation Climatology Project version 2.1 (GPCP; Adler et al. 2003), suggest that 2013 may have been somewhat below average, though GPCP has generally been slightly drier than the other products in recent years.
Several coherent anomaly patterns were evident over land in 2013 (Plate 2.1h). For instance, below-average precipitation fell over much of North America, northern Eurasia, southern South America, sub-Saharan Africa, and Australia. In contrast, above-average precipitation fell over parts of southern Asia, the Amazon basin, the Maritime Continent, and Greenland (though the latter should be viewed cautiously because of the statistically in-filled estimates in that area). Relative to 2012, the dry conditions over central North America and eastern South America became somewhat less extreme. Meanwhile, much of the Sahel flipped from above- to below-normal rainfall.
Globally, precipitation over the oceans was above the 1988-2010 average in 2013 (Fig. 2.15b,c). This conclusion is based on intercalibrated passive microwave retrievals in the Remote Sensing Sytems (RSS; Hilburn and Wentz 2008) version 7 product, which was about 12 mm above normal. Ocean data for two other products, the GPCP blended satellite-in situ dataset and the Climate Prediction Center Merged Analysis of Precipitation (CMAP; Xie and Arkin 1997), are not yet available for 2013. Relative to terrestrial datasets, the ocean products are less similar globally, though some consistency is evident--e.g., the wettest recent year was 2010,2011 was much drier, and ocean precipitation has increased since then. RSS has been wetter than GPCP, which has been wetter than CMAP.
Coherent anomaly patterns were evident over the ocean as well in 2013 (Plate 2.1h). For instance, there were dry anomalies in much of the Indian Ocean and the Arabian Sea. Wet anomalies extended southward from the Bay of Bengal, along Sumatra, and between Australia and Indonesia, covering the Timor and Arafura Seas. The wet anomalies extended northward through the Java and Banda Seas, as far north as the South China Sea, and as far east as the Caroline Islands. The North Pacific had wet anomalies from Japan to Alaska, and the South Pacific had wet anomalies along the South Pacific convergence zone. It was drier over the North Atlantic, especially along North America, extending as far south as the Caribbean Sea.
4) CLOUDINESS--M. Foster, S. A. Ackerman, A. K. Heidinger, B. C. Maddux, and M. Stengel
Global mean annual cloudiness anomalies from six satellite data records, a synoptic record, and a reanalysis product are shown in Fig. 2.16. The PATMOS-x (Pathfinder Atmospheres Extended) and MISR (Multiangle Imaging Spectroradiometer) records show mean global cloudiness remained static from 2012 to 2013 (within 0.1%) while the Moderate Resolution Imaging Spectroradiometer (MODIS) showed a moderate increase of 0.4%. ISCCP (International Satellite Cloud Climatology Project), HIRS (High Resolution Infrared Sounder), and CLARA-A1 (Cloud, ALbedo and RAdiation dataset) are also shown though they currently do not extend through 2013. CLARA-A1 is a EUMETSAT Satellite Application Facility on Climate Monitoring (CM-SAF) climate application product consisting of cloud, surface albedo, and radiative parameters. It covers 1982-2009 with plans for future reprocessing. The synoptic record, SYNOP, consists of manual cloud cover observations in oktas. For inclusion SYNOP sites must have available observations for at least 75% of the record (1981-2013). Mainly North American and European sites fulfill this criterion, so it is not a true global record. Cloudiness is also provided from the ERA-Interim reanalysis.
Historically, 2013 was the sixth least cloudy year, 1.1% less cloudy than the mean for the 33-year PATMOS-x record, the primary data set used here. There is general agreement among the satellite records, although this deteriorates away from the common reference period of2000-09, especially for CLARA-A1. However, there is comparative interannual stability in mean global cloudiness since 2000. Variability in the early part of the CLARA-A1 record may in part be attributed to the combined effect of imperfect diurnal sampling and satellite drift. Cloud detection over semi-arid areas during daytime also contributes and is more pronounced during the 1980s when no morning satellites were used. In the case of PATMOS-x, which like CLARA-A1 is derived from the AVHRR (Advanced Very High Resolution Radiometer) sensor, a diurnal correction has been applied to account for this issue (Foster and Heidinger 2013). There is no consensus in any long-term trends among the records.
Global cloudiness in 2013 was characterized by a lack of any strong ENSO signal and an almost complete lack of statistically significant positive anomaly regions (Plate 2.Id; Online Fig. S2.13). Accordingly, there were few maritime cloudiness anomalies over the tropical and subtropical Pacific, as seen in Plate 2.Id where the cloudiness anomaly is defined as the 2013 cloudiness minus the climatological mean cloudiness (1981-2010, using PATMOS-x). Negative anomalies occurred off the northern coasts of Alaska and Scandinavia, consistent with the negative phase of the Arctic Oscillation present early in the year (section 2el). But, although the negative annual anomalies are significant at the 5% level, the Arctic Oscillation shifted to a positive phase in the middle of the year, contributing to some of the wettest months on record for Alaska and Norway.
Several significant (at the 5% level) negative continental anomalies correspond to severe drought (compare Plate 2.1h) or heat extremes. In the western United States and southern Africa (including Madagascar) these anomalies corresponded with severe drought conditions. In southern China, South Korea, and Japan the reduced cloudiness corresponded with heat extremes in the summer months, while in Russia warm extremes were observed in the winter months (see Online Fig. S2.13; sections 7b, d, g).
ENSO affects the global distribution of cloudiness through enhancement/suppression of large-scale convection in the western Pacific and corresponding suppression/enhancement of convection in the central Pacific driven by SST and low-level wind gradients between the central equatorial Pacific and Indonesia. While the historical ENSO signal can be seen in the Hovmoller for 1981-2013 in Online Fig. 2.14, it is not clear how consistently the geographic distribution of mean annual cloudiness relates to the strength of ENSO alone. To this end, the annual mean total cloudiness anomalies for each year in the PATMOS-x record have been calculated here. The geographic region between 70[degrees]N and 70[degrees]S is considered, as the lack of radiometric contrast between clouds and the surface in polar regions introduces significant uncertainty into the detection process. A clustering algorithm is applied (Anderberg 1973), which sorts those years with similar spatial patterns of cloudiness into distinct groups or 'clusters'. Three clusters were created and when plotted annually along with a multivariate ENSO index (Wolter and Timlin 1998; Fig. 2.17) it is clear that ENSO is a primary driver of cloudiness, as cluster membership corresponds closely with index value. Each cluster can be thought of as a "climate state" (CS). In this context, CS1 represents La Nina conditions and corresponds to negative cloudiness anomalies over the central equatorial and southeastern Pacific, CS2 represents ENSO-neutral conditions with neutral cloudiness, and CS3 represents El Nino conditions with positive cloudiness anomalies over the central equatorial and southeastern Pacific. The year 2013 was determined to fit into CS2, with relatively average cloudiness conditions. The last time a year was classified as CS3 was 1998, before the start of MODIS and MISR record, though it should be noted that increasing the number of clusters allows for more refined ENSO states. For example when five clusters are created one represents weak El Nino conditions with positive cloudiness anomalies concentrated in the central Pacific, to which 2002 would belong.
5) RIVER DISCHARGE--H. Kim and T. Oki
As an integral connection between overland precipitation and freshwater supply to oceans, river discharge is an essential component in the global water cycle. Its availability has affected civilizations considerably through cultural and economic impacts over time. However, due to the absence of direct observations at the global scale and the decreasing number of gauge stations (Fekete et al. 2012), off-line hydrologic modeling is still required to estimate river discharge at long-term global scales. Global discharge excluding Greenland and Antarctica is estimated as a subset of the Ensemble Land Surface Estimator (ELSE; Kim et al. 2009). The Japanese 25-year Reanalysis (JRA-25; Onogi et al. 2007) provides surface meteorological variables from 1979 through 2013 to force the land surface model Minimal Advanced Treatments of Surface Interaction and RunOff (MATSIRO; Takata et al. 2003). Realistic month-to-month variability is introduced using the Global Precipitation Climatology Project Version 2.2 (GPCP; Huffman et al. 2012) and Global Precipitation Climatology Centre (GPCC; Rudolf and Rubel 2005) monthly observational precipitation products. Due to the lagged update frequency of the GPCC Full Data Reanalysis Version 6, the Monitoring product Version 4 is used for the period of 2011-13. In order to reconcile the time series of these two separated periods, a trend-preserving statistical bias correction (Watanabe et al. 2012) is applied. In addition, a wind-induced under-catch correction (Legates and Willmott 1990) is applied to GPCC precipitation estimates. Simulated runoff is routed through a global river transfer model, Total Runoff Integrated Pathway (TRIP; Oki and Sud 1998). Simulations are validated over 29 global river basins which encompass approximately 25% (32 358 232 [km.sup.2]) of the global terrestrial area (130764683 [km.sup.2]). Both flux (discharge) and storage (terrestrial water storage) terms are compared against Global Runoff Data Center in-situ observations and the Gravity Recovery And Climate Experiment (GRACE; Tapley et al. 2004) satellite remote sensing data, respectively (http://hydro.iis.u-tokyo.ac.jp/~hjkim/tws@2009GRL/).
Plates 2.1i and 2.1j show spatial variability of the global river discharge and runoff anomaly in 2013, and Fig. 2.18 shows continent-wise runoff anomaly estimations during the recent four years. Strong spatial variability is apparent during 2013. Within South America, the northwestern part of the continent and most of the Amazon basin show wetter conditions than the climatological mean (1979-2013) in contrast to the drier conditions of the southeastern parts such as Rio de la Plata and Tocantins. In the annual analysis, most of North America suffers from drier conditions than normal, except in the southeastern part of the United States. During the last four years runoff from North America tends to be below the long-term mean, and 2013 is the second driest year following 2012. While the Nile River has more freshwater discharge, the other major river basins on the African continent (e.g., Congo, Niger, Zambezi, and Orange) show less discharge in 2013. The annual discharge from this continent is consistently below the average for the last few years. Relatively weak interannual variability is found in the recent annual estimates of discharge from the European continent. Mediterranean countries are wetter on average while northern European countries are drier leading to a neutral continental balance. The high latitudes of the Eurasia continent and East Asia show negative anomalies particularly for the Yenisei River and its vicinity and rivers in China. The Ob, Amur, and Brahmaputra Rivers transport more water than average. Australia shows extremely large interannual variability. The discharge in 2013 shows a significant negative relative anomaly in contrast to the extreme positive anomaly in 2011.
6) GROUNDWATER AND TERRESTRIAL WATER STORAGE--M. Rodell, D. P. Chambers, and J. S. Famiglietti
Terrestrial water storage (TWS) comprises groundwater, soil moisture, surface water, snow, and ice. Groundwater typically varies more slowly than the other TWS components because it is not in direct contact with the atmosphere; however, it often has a larger range of variability on multiannual timescales (Rodell and Famiglietti 2001; Alley et al. 2002). In situ groundwater data are only archived and made available by a few countries. However, monthly TWS variations observed by the Gravity Recovery and Climate Experiment (GRACE; Tapley et al. 2004) satellite mission, which launched in 2002, are a reasonable proxy for unconfined groundwater at climatic scales.
Changes in mean annual TWS from 2012 to 2013 are plotted in Plate 2.1k as equivalent depths of water in cm. TWS can be thought of as an integrator of other hydroclimatic variables (see Plates 2.1d-2.11). Many parts of the Northern Hemisphere saw a recovery in 2013 from the dry conditions of 2012, while drought continued in other areas. The massive drought that covered most of North America in 2012 abated in much of the eastern and central United States and Canada, but worsened to near-record levels in the southwestern United States. Europe and Russia also recovered from a dry 2012. The year was mixed in southern Asia, with drought afflicting Bangladesh and eastern and southern India. Depletion of aquifers by pumping for irrigation continued in northern India (Rodell et al. 2009; Tiwari et al. 2009) and the North China Plain (Feng et al. 2013), while heavy rains in parts of Turkey and the Middle East helped raise otherwise depressed water levels (Voss et al. 2013). Parts of southern Africa, including Angola and Namibia, went from moderately dry in 2012 to severe drought in 2013. In South America, the central Amazon became extremely wet, while parts of coastal Brazil and Venezuela were dry for most of the year. Australia as a whole lost a large amount of TWS in 2013. Significant reductions in TWS in Greenland, Antarctica, and southern coastal Alaska represent ongoing ice sheet and glacier ablation, not groundwater depletion.
Figures 2.19 and 2.20 show time series of zonal mean and global, deseasonalized monthly TWS anomalies from GRACE, excluding Greenland and Antarctica. Data gaps occur when the satellites were powered down to conserve battery life. Recovery from the unusually dry conditions of 2012 can be seen, particularly in the northern midlatitudes (Fig. 2.19), and also in the global land (Fig. 2.20). The global TWS anomaly ended 2012 at -15 cm, reached an 11-year minimum of-18 cm in February 2013, and recovered to -2 cm by December 2013.
7) SOIL MOISTURE--W. A. Dorigo, D. Chung, R. M. Parinussa, C. Reimer, S. Hahn, Y. Y. Liu, W. Wagner, R. A. M. de Jeu, C. Paulik, and G. Wang
Soil moisture is both a manifestation and a driver of the complex interactions between the water, energy, and biogeochemical cycles at the Earth's surface (e.g., Taylor et al. 2012). Monitoring long-term changes in its mean values and variability is thus pivotal for understanding the effects of climate change (Seneviratne et al. 2010). In 2012 the Climate Change Initiative (CCI) of the European Space Agency released a soil moisture dataset (ECV soil moisture) that amalgamates global observations from various space-borne radiometers and scatterometers (De Jeu et al. 2012a; Liu et al. 2012). Recently, the ECV soil moisture product has undergone several algorithmic improvements and has been complemented with observations from the Coriolis Windsat and GCOM-W AMSR2 sensors to continue the legacy of C-band observations in the passive microwave domain. The observation record now spans a 35-year period (late 1978-present). Anomalies are based on a 1991-2012 climatology. The first 13 years contain different dataset characteristics and so are not included in the climatology.
Plate 2.11 shows where in 2013 either dry (brown) or wet (blue) anomalous conditions prevailed. Anomalous dry conditions were observed in particular in the Southern Hemisphere, e.g., in Argentina, northeastern Brazil, southern Africa, and Australia. These areas are particularly sensitive to drought during the El Nino phase of ENSO (Bauer-Marschallinger et al. 2013; Miralles et al. 2014). However, ENSO conditions were neutral throughout 2013 (see section 2el). For some of the drought-affected regions (e.g., northeastern Brazil and southern Africa) strong anomalous negative soil moisture conditions were present in 2012 (Parinussa et al. 2013) and continued in 2013. The monthly anomaly maps (Online Fig. S2.15) show that negative anomalies were particularly evident during the first half of 2013, but gradually decreased towards the end of the year, except for Australia where heat records throughout the year led to an annual high temperature record (BoM 2014; also see Sidebar 7.5). Dry conditions were also observed in southeastern Europe in summer, caused by a lack of precipitation and high temperatures (see section 7f). After several extremely dry summers in a row (De Jeu et al. 2011, 2012b; Parinussa et al. 2013), central Eurasia experienced neutral conditions in 2013 (Online Fig. S2.15). Similarly, in the southeast and central parts of the United States, after several very dry years, drought conditions recovered towards the end of summer; however, conditions deteriorated in the far west in the course of 2013, with record to near-record dryness (see section 7b2). This evolution is visible in the monthly maps (Online Fig. S2.15).
Extremely wet conditions during boreal summer likely prompted several major flood events such as those that occurred in the eastern parts of Russia and northeast China in August 2013. Anomalously wet conditions were also observed in March in Spain, which received more than triple its average precipitation for this month. Extremely wet conditions in South Asia (June, July, and October) and Brazil (December) contributed to floods and landslides.
On a global scale, 2013 soil moisture did not strongly deviate from the last few years (Fig. 2.21). This is mainly attributed to the absence of a strong ENSO anomaly which drives the major year-to-year variability in average terrestrial wetness conditions. ENSO-driven global variations are particularly visible during the El Nino of 1997-98 and the La Nina of 2010-11, and closely correspond to the dynamics observed in terrestrial water storage from GRACE (Boening et al. 2012). When looking at the entire time series, average soil moisture conditions over 2013 seem to confirm the slight general drying trend as observed by Dorigo et al. (2012), but are not in line with recent trends in global evaporation (Miralles et al. 2014) or drought severity (van der Schrier et al. 2013). These differences between products are due to various factors, including the physical mechanisms and forcing of the latter two products and the incomplete global coverage of the ECV soil moisture product, which is masked for dense vegetation and frozen soil conditions (see also Fig. 2.21).
The year-to-year and seasonal variability is even more apparent at regional scales (Fig. 2.22). For example, for the southern midlatitudes, an alternation of dry and wet periods can be observed, including the 2001-09 Australian Millennium Drought (van Dijk et al. 2013), followed by extremely wet conditions in 2010-11, and turning back to drought conditions again during the last two years. The figure also illustrates that the general drying trend observed for the Northern Hemisphere over the last decade is observed across the entire midlatitudes, including the southern United States and the Mediterranean.
e. Atmospheric circulation
1) MEAN SEA LEVEL PRESSURE--R. Allan and C. K. Folland
El Nino and La Nina can be measured by the Southern Oscillation index (SOI), the normalized MSLP difference between Tahiti and Darwin (Allan et al. 1996). It can also be measured using sea surface temperatures (SSTs; e.g., see Fig. 4.1). El Ninos (negative SOI) and La Ninas (positive SOI) vary in magnitude, duration, and evolution, with no two events exactly the same. Major El Nino and La Nina events can be near-global in their influence on world weather patterns, owing to ocean-atmosphere interactions across the Indo-Pacific region with teleconnections to higher latitudes in both hemispheres.
Since the termination of the protracted La Nina episode in early 2012, near-normal SOI values persisted until early 2013, after which values were mainly positive for the rest of the year (Fig. 2.23a,b). However, these values did not reach the threshold for an official La Nina event, and overall 2013 was ENSO-neutral.
The SOI is arguably the most global mode of sea level pressure variability. Other regionally notable modes are shown in Fig. 2.23. In late 2012, positive North Atlantic Oscillation (NAO; see Fig. 2.24a,c)/ Arctic Oscillation (AO) conditions favored a westerly regime and wet weather over Europe, with major flooding episodes in the United Kingdom. But in early 2013 this westerly circulation pattern was replaced by more easterly winds over western Europe and the United Kingdom, leading to colder, snowy weather (see section 7f for more details).
In contrast, the Northern Hemisphere winter of 2013/14 experienced a different mix of conditions. Since December 2012, the North Pacific anticyclone has been anomalously strong and persistent, leading to prolonged drought in California (Plates 2.1d-2.1m). This strong anticyclone (Fig. 2.24a,b) was accompanied in the Northern Hemisphere early winter of 2013/14 by a positive AO, a deep trough over Canada and the United States, and a southerly displaced and enhanced subtropical jet-stream extending across the Atlantic to the United Kingdom and Europe under strong positive NAO conditions (Fig. 2.24d). This led to severe cold winter conditions in much of the United States and a succession of major midlatitude storms being steered across the Atlantic to Ireland and the United Kingdom.
In the Southern Hemisphere, the Antarctic Oscillation (AAO) did not exhibit strong features during the austral summers of 2012/13 or early 2013/14 (Fig. 2.23e,f).
In the main boreal summer months of July and August, the summer NAO (SNAO), which has a northward displaced pattern compared to the winter NAO (Folland et al. 2009), behaved differently from the generally negative phase observed during the high summers of 2007-12. Figure 2.25a shows daily SNAO values expressed as anomalies from 1981-2010 for July and August 2013 compared to the daily average for 2007-12. The strongly positive phase in 2013 was the first since 2006, except for a marginally above-zero value in 2010, reflected in the pattern of North Atlantic and Europe MSLP anomalies for high summer 2013 (Fig. 2.25b). The pattern is much like that of the positive phase the SNAO in Folland et al. (2009). The strong anticyclonic anomaly over northwest Europe was associated with generally warm and dry conditions (Online Figs. S2.16, S2.17) so that 2013 was the 14th warmest high summer in the central England temperature record back to 1659. England and Wales rainfall was correspondingly dry at 84% of average, the first dry July and August since 2006 (see section 7f for more details).
2) SURFACE WINDS
(i) Land surface wind speed--I. Tobin, P. Berrisford, R. J. H. Dunn, R. Vautard, and T. R. McVicar
Surface wind over land is observed at weather stations using anemometers a few meters above the ground. Surface wind speed can vary rapidly over time and space, and station networks are irregularly distributed and sparse, especially in the Southern Hemisphere (see station positions in Plate 2.In; Fig. 2.26), leading to concerns about representativeness. Following McVicar et al. (2013), three strictly quality controlled datasets are used here: reduced ISD-Lite (Vautard et al. 2010), HadlSD (Dunn et al. 2012), mainly over the Northern Hemisphere, and an Australian database (McVicar et al. 2008) made up of about 1350, 2500, and 40 stations, respectively. Reanalysis products provide contiguous global information (ERA-Interim is used here, Dee et al. 2011; Online Fig. S2.18) but exhibit shortcomings in capturing surface winds, as many surface-layer processes controlling wind are not adequately represented (McVicar et al. 2008; Pryor et al. 2009; Vautard et al. 2010). Years prior to 1981 suffer from a significant lack of records in the ISD-Lite database and are thus not considered.
While still below the long-term climatology, 2013 overall winds recorded at stations represent a small increase (-0.05 m [s.sup.-1]) compared to the preceding three years (Fig. 2.27a). The exception is North America where 2013 wind speed was the fourth lowest on record since 1981. Over Europe 2013 was the 10th (12th) lowest year according to ISD-Lite (HadISD), and it was also the 10th lowest year over Australia. The year 2013 was an average year over central and eastern Asia, which exhibit slightly negative anomalies in ISD-Lite and small positive anomalies in HadISD. Figures 2.27b and c show that this short-term strengthening of mean wind speed relative to previous years was made up of an increase in frequency of low and moderate winds (>3 m [s.sup.-1]), especially over Asia. However, it was not accompanied by an increase in strong wind (>10 m [s.sup.-1]) frequency which is similar to previous years over all the regions. Both categories remain below the 1981-2010 climatological average, by 2% and 0.02%, respectively, according to ISD-Lite.
Spatial patterns of anomalies are shown in Plate 2.In and Online Fig. S2.18. The magnitude of anomalies does not exceed 0.5-1 m [s.sup.-1] for most stations. North America is dominated by negative anomalies while other regions are characterized by both negative and positive anomalies. The large-scale anomaly patterns shown by ERA-Interim are consistent with station data and the anomaly magnitude is overall reasonably reproduced despite the substantial spatial resolution difference between datasets (point vs -0.7[degrees] x 0.7[degrees] grid). Over northern and southern Africa and Saudi Arabia ERA-Interim shows extended positive anomalies; however, there are no stations located in these regions from ISD-Lite or HadISD with which to compare. Likewise, South America is characterized by positive anomalies in ERA-Interim, similar to the previous three years (Peterson et al. 2011; Vautard et al. 2012; McVicar et al. 2013).
With regard to the northern midlatitudes and Australia, 2013 is embedded in a 33-year stilling trend, ranging on average from about 0.07 to 0.1 m s4 decade-1 (Table 2.5; Fig. 2.26a,b). The slowdown of land surface winds has already been reported over many regions (see McVicar et al. 2012 for a review; Dadaser-Celikand and Cengiz 2013; C. Lin et al. 2013; Azorin-Molina et al. 2014). In 2013 this tendency continued over North America. The short-term strengthening of winds observed over other regions only slightly affects the trend assessment (75% of ISD-Lite stations considered here exhibit trend assessment changes of less than 15% compared with 2012; Fig. 2.26). Although the ERA-Interim pattern of trends is consistent with station data, the magnitude is significantly underestimated as previously noted with other reanalysis products (McVicar et al. 2008; Pryor et al. 2009; Vautard et al. 2010).
This stilling is not fully understood and does not necessarily reflect wind tendency at higher levels (Vautard et al. 2010; Troccoli et al. 2012). Vegetation cover increase, air pollution, thermal and pressure gradient decrease, and urbanization are among the identified causes, which differ among regions (Dadaser-Celik and Cengiz 2013; C. Lin et al. 2013; Azorin-Molina et al. 2014; McVicar et al. 2013).
(ii) Ocean surface wind speed--C. Mears
Estimates of globally-averaged wind over the oceans obtained from satellite-borne microwave radiometers were slightly lower than average for 2013 (Wentz 1997; Wentz et al. 2007; Fig. 2.28; Online Fig. S2.19). Estimates from reanalysis products differ with JRA-55 and ERA-Interim showing that 2013 was windier than normal, and MERRA showing the opposite. Reanalysis winds, which are in relatively good agreement with both the satellite data and each other from 1990-2009, diverge after 2010. The in situ data show larger trends than the satellite datasets. When comparing global means between in situ datasets and satellite datasets, it is important to note that the area coverage of the in situ datasets is less, with areas of missing or poorly sampled data in the tropical Pacific and the Southern Ocean. The newer JRA-55 product is in better agreement with the satellite measurements than the previous JRA-25 product, which showed a larger increasing trend during the satellite period than JRA-55. All products show an increasing trend from 1990 to 2007, followed by a drop-off in 2008-09, and a recovery in 2010. Since then the winds have slackened.
During 2013, winds showed positive anomalies in the central tropical Pacific (Plate 2.In). These anomalies were present and much larger during the 2010-12 period. Other regions with positive anomalies include off the eastern coast of North America, west of Spain, the Mediterranean Sea, the northern Indian Ocean, and the Pacific Ocean south of about 50[degrees]S. The Gulf of Alaska showed a large negative anomaly in association with persistent high pressure (Plate 2.1m and section 2el).
A trend map of wind speed over the satellite era (1988-2013; Fig. 2.29a) shows a region of increasing windiness in the central tropical Pacific. These winds typically blow toward the west, leading to increases sea surface height in the western Pacific that are strongly correlated with the wind anomalies (Fig. 2.29b).
f. Earth radiation budget
I) EARTH RADIATION BUDGET AT TOP-OF-ATMOSPHERE--D. P. Kratz, P. W. Stackhouse, Jr., I. Wong, P. Sawaengphokhai, A. C. Wilber, S. K. Gupta, and N. G. Loeb
The Earth's radiation budget (ERB) at the topof-atmosphere (TOA) is defined as the sum of the incoming total solar irradiance (TSI), the reflected shortwave radiation (RSW), and the outgoing longwave radiation (OLR). Since the relationship between the incoming and outgoing energies defines the climate state of the Earth-atmosphere system, quantifying these values is of utmost importance in understanding the energy budget that drives weather processes, climate forcing, and climate feedbacks.
An analysis of all measurements from 2012-13 (Table 2.6) shows that the global annual mean OLR increased by ~0.25 W [m.sup.-2] and the RSW increased by -0.45 W [m.sup.2]. Over the same timeframe, the TSI remained essentially constant. The sum of these components amounts to a reduction of-0.70 W [m.sup.-2] in the total net radiation into the Earth climate system for 2013 compared with 2012. Relative to the multiyear data average (2001-12), the 2013 global-annual mean flux anomalies (Table 2.6) are -0.05, +0.05, +0.20, and -0.10 W [m.sup.-2] for OLR, TSI, RSW, and total net flux, respectively. These changes are within the corresponding 2-sigma interannual variability for this period.
Prior to August 2013, the TSI data were obtained from either the Total Irradiance Monitor (TIM) instrument aboard the Solar Radiation and Climate Experiment (SORCE) spacecraft or other satellite data renormalized to SORCE (Kopp and Lean 2011). After the SORCE spacecraft's battery failure in July 2013, the TSI data were obtained from the Royal Meteorological Institute of Belgium (RMIB) composite dataset (Dewitte et al. 2004). To merge the SORCE and RMIB datasets, a time-dependent scaling factor was used to calibrate the differences between the two datasets from March 2003 through June 2013. The RMIB data was then used to simulate SORCE TSI values for July-December 2013. The RSW and OLR data were obtained from the Clouds and the Earth's Radiant Energy System (CERES) mission (Wielicki et al. 1998), which has been deriving flux data from the CERES measurements taken aboard the Terra and Aqua spacecraft since March 2000 and July 2002, respectively. This report focuses on the most recent measurements relative to the long-term CERES dataset.
The monthly mean anomaly time series for the TOA flux components covering March 2000-December 2013 are presented in Fig. 2.30. The OLR oscillated between [+ or -] 0.60 W [m.sup.-2] throughout 2013 before reaching a neutral value of -0.04 W [m.sup.-2] in December 2013. The observed OLR variability is generally consistent with the Atmospheric Infrared Sounder (AIRS) OLR data (monthly AIRX3STM.006 product; not shown). A comparison of the OLR to the multivariate ENSO index (not shown) revealed no trend in the OLR, consistent with the relatively neutral ENSO condition persisting throughout 2013. The absorbed shortwave (TSI - RSW) fluctuated between -0.81 and [+ or -] 0.97 W [m.sup.-2] during 2013, ending the year at the maximum. The total net anomaly, which contains the combined OLR and absorbed shortwave anomalies, began 2013 with a pronounced minimum of -1.24 W [m.sup.-2], then oscillated between [+ or -] 0.60 W [m.sup.-2] for most of the year before rising to a maximum.of +1.02 W nr2 at the end of 2013.
Temporal analysis from March 2000 to December 2013 was achieved through the merger of two ERB datasets: (1) the CERES EBAF (Energy Balanced And Filled) Ed2.7 product (Loeb et al. 2009, 2012), March 2000-June 2013, and (2) the CERES FLASHFlux (Fast Longwave and Shortwave Radiative Fluxes) 3A product (Kratz et al. 2014; Stackhouse et al. 2006), July-December 2013. The FLASFIFlux components are normalized to the EBAF Ed2.7 data using TOA fluxes from both datasets for the 1-year overlap period from July 2012-June 2013. The EBAF data products use TSI from the SORCE mission while the FLASFIFlux data products assume a constant TSI, modified only for earth-sun distance. The FLASHFlux TSI data were then adjusted to the combined SORCE and renormalized RMIB data and the RSW was scaled accordingly. The resulting 2-sigma monthly uncertainty of the normalization procedure for the 1-year overlap period was [+ or -] 0.34, [+ or -] 0.05, [+ or -] 0.84, and [+ or -] 0.93 W [m.sup.-2] for the OLR, TSI, RSW, and NET radiation, respectively. The normalization coefficients were then applied to the FLASHFlux data to obtain an EBAF-compatible TOA radiative flux time series through the end of 2013. Owing to the observed variability in the six months extending beyond the EBAF data, long-term trend analyses with the merged data set are discouraged due to the natural fluctuation in ERB components, the uncertainty from the data merging process, and potential for drift in the FLASHFlux product.
2) MAUNA LOA CLEAR-SKY ATMOSPHERIC SOLAR TRANSMISSION--K. Lantz
Atmospheric solar transmission has been measured for five and a half decades by the Global Monitoring Division (GMD) of the National Atmospheric and Oceanic Administration (NOAA) at the Mauna Loa Observatory (MLO). Mauna Loa Observatory is at 3400 m elevation on the northern slope of the Mauna Loa volcano on the Big Island of Hawaii. The remote location and high altitude make it well-suited for studying changes in the free troposphere with limited local influences. A clear-sky "apparent" solar transmission (AT) is calculated using the ratio of direct-beam broadband pyrheliometer measurements at fixed atmospheric paths (air mass; Ellis and Pueschel 1971). The AT is advantageous because it is independent of the radiometer calibration and the extraterrestrial irradiance. The MLO AT is particularly sensitive to changes in background stratospheric aerosols and the influence of volcanic eruptions. Studies have examined the variability in the MLO AT due to water vapor, ozone, volcanic aerosol, aerosol transport from Asia, and atmospheric circulation changes associated with the quasi-biennial oscillation (QBO; Bodhaine et al. 1981; Dutton et al. 1985, Dutton 1992; Dutton and Bodhaine, 2001).
The updated clear-sky AT from the 1958-2013 monthly record is computed from daily early morning values to remove boundary layer influences due to upslope winds (Fig. 2.31a). The aerosol signal from the eruptions of Agung, El Chichon, and Mount Pinatubo in 1964,1982, and 1991, respectively, are clearly visible in the record. The 6-month running smoothed fit to the monthly values highlights the seasonal trends in the data that have been attributed primarily to Asian aerosol transport in the spring (Bodhaine et al. 1981). This seasonal variability of the clear-sky AT has an amplitude of 0.007. A 24-month running smoothed fit highlights the longer-term changes. The gray dashed line reflects the cleanest background observed from 1958 to 1962 in the record, except for a brief period in 1978. The average clear-sky AT in 2013 has increased with respect to 2012 but is still not as clean as the cleanest background observed between 1958 and 1962.
Annual clear-sky AT averages of the 10 cleanest days are useful for viewing stratospheric background air while limiting the influence of local air pollution events or contamination by cirrus clouds. Previous studies showed that the annual clear-sky AT returned to near-background conditions after the eruption of Mount Pinatubo with a subsequent slow decrease in the clear-sky AT that was in concert with a slow increase in the background annual average MLO AOD (PFR) both based on the 10 cleanest days in the year from the years 2000 to 2010 (Solomon et al. 2011). This decrease in AT from 2000-10 is also clearly evident in the 24-month running smoothed fit in Fig. 2.31b. Solomon et al. (2011) showed that four independent data-sets confirmed an increasing background stratospheric aerosol from 2000-10 (i.e., clear-sky AT at MLO, PFR AOD at MLO, stratospheric AOD from lidar at MLO, and AOD from tropical satellite measurements above 15 km). This was attributed to possible smaller volcanic eruptions since 2000 (Vernier et al. 2011). This changing background stratospheric aerosol was shown to have implications for climate change and changes in surface temperature (Solomon et al. 2011). Earlier work suggested a possible leveling of this decreasing AT after 2010 (Dutton 2012). The additional years in the MLO AT record continue to show a "persistently variable background" AT but it is no longer decreasing and shows leveling in the last few years (Fig. 2.31b).
g. Atmospheric composition
I) LONG-LIVED GREENHOUSE GASES--E. J. Dlugokencky, B. 0. Hall, S. A. Montzka, G. Dutton, J. Miihle, and J. W. Elkins
Carbon dioxide (C02) is the dominant long-lived greenhouse gas (LLGHG) contributing to climate forcing; since 1750 its radiative forcing has increased by 1.88 W m 2 or -65% of the increased forcing by all LLGHGs (see http://www.esrl.noaa.gov/gmd/aggi/aggi.html). When systematic C[O.sub.2] measurements began at Mauna Loa, Hawaii, (MLO) in 1958, the annual mean mole fraction was -315 parts per million (ppm). In May 2013 daily-averaged C[O.sub.2] at MLO exceeded 400 ppm for the first time (see http://www.esrl.noaa.gov/gmd/ccgg/trends/index.html). This 27% increase is mainly due to a fourfold rise in anthropogenic C[O.sub.2] emissions from fossil fuel combustion and cement production. The C[O.sub.2] growth rate has correspondingly increased from 0.7 ppm yr1 in the early 1960s to 2.1 ppm yr1 during the last decade. About half of the C02 emitted remains in the atmosphere; the rest is taken up by the oceans and terrestrial biosphere. The annual atmospheric increase varies considerably from year to year, ranging from 0.7 [+ or -] 0.1 to 2.8 [+ or -] 0.1 ppm [yr.sup.-1] since 1990. This is explained largely by variations in natural fluxes influenced by the phase of ENSO (Bastos et al. 2013). In 2013 the globally averaged CO, mole fraction at Earth's surface was 395.3 [+ or -] 0.1 ppm (Fig. 2.32a), an increase of 2.8 [+ or -] 0.1 ppm over the 2012 mean.
Atmospheric methane (C[H.sub.4]) has contributed -0.5 W [m.sup.-2] direct radiative forcing since 1750. Indirect effects from the production of tropospheric ozone (03) and stratospheric water ([H.sub.2]O) added another -0.2 W [m.sup.-2]. Atmospheric methane is produced by natural (-40%) and anthropogenic (-60%) sources. Natural sources include wetlands, geological sources, oceans, and termites (Dlugokencky et al. 2011). Anthropogenic sources include agriculture (ruminants, rice), fossil fuel extraction and use, biomass burning, landfills, and waste. Fossil fuel exploitation (coal, oil, and natural gas) contributes -20% of total global C[H.sub.4] emissions (Kirschke et al. 2013). New methods of oil and gas extraction may emit large amounts of C[H.sub.4] locally (e.g., Karion et al. 2013), but these emissions are currently small relative to the anthropogenic total. Atmospheric C[H.sub.4] has increased by about a factor of 2.5 since the pre-industrial era (1750). The annual rate of increase of >10 parts per billion (ppb) [yr.sup.-1] in the 1980s slowed dramatically to near zero in the early 2000s, then in 2007 jumped to ~6 ppb [yr.sup.-1] and has remained fairly steady since (Fig. 2.32b). Global emissions are estimated to be -540 Tg C[H.sub.4] [yr.sup.-1] ([+ or -] 10%) based on observations of the global C[H.sub.4] burden and an estimate of its atmospheric lifetime (-9.1 yr). This top-down emission estimate provides an important constraint on bottom-up inventories of C[H.sub.4] emissions that tend to overestimate the global total (Kirschke et al. 2013). The same observation-based mass-balance estimation of global emissions can also be applied to most other LLGHGs. Based on NOAA background air sampling sites, the globally averaged C[H.sub.4] mole fraction at Earth's surface in 2013 was 1814.1 [+ or -] 0.8 ppb, a 5.7 [+ or -] 0.9 ppb increase from 2012 that conforms to the average growth rate since 2007.
Nitrous oxide ([N.sub.2]O) currently exerts the third strongest climate forcing of the LLGHGs after C[O.sub.2] and C[H.sub.4]. Nitrous oxide is produced in soils by both the oxidation of ammonium and the denitrification of nitrate. About one-third of [N.sub.2]O emissions are related to human activities, and about two-thirds of these are related to the agricultural application of nitrogen-containing fertilizers, including manure (UNEP 2013; Davidson 2009; Reay et al. 2012). The mean global atmospheric [N.sub.2]O mole fraction in 2013 was 325.9 ppb, an increase of 0.9 ppb from 2012 (Fig. 2.32c). This growth rate exceeds the 2000-13 average of 0.78 [+ or -] 0.01 ppb [yr.sup.-1] and is consistent with the more recent 2010-13 mean growth rate of 0.92 [+ or -] 0.02 ppb [yr.sup.-1].
Halogenated gases, such as chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), and hydrofluorocarbons (HFCs), also contribute to radiative forcing. While the atmospheric mole fractions of some CFCs, such as CFC-12 and CFC-11 are decreasing (Fig. 2.32d), the atmospheric burdens of their industrial replacements HCFC-22 and HFC134a are increasing (Fig. 2.33). After a period of enhanced growth in HCFC-22 that began around 2005 (Montzka et al. 2009) the growth rate of HCFC-22 in the atmosphere slowed slightly in recent years. The abundance of sulfur hexafluoride (SF6), used in high-voltage electrical transmission equipment, also continues to increase. The mean global SF6 mole fraction in 2013 was 7.91 parts per trillion (ppt), 0.32 ppt higher than in 2012. Global annual mean mole fractions of a number of trace gases, along with their 2012 to 2013 changes are listed in Table 2.7.
Recent trends in the combined radiative forcing by five major LLGHGs (C[O.sub.2], C[H.sub.4], N.sub.2]O, CFC-11, and CFC-12) and 15 minor gases are illustrated by the NOAA Annual Greenhouse Gas Index (AGGI; Hofmann et al. 2006; http://www.esrl.noaa.gov/gmd/aggi/). This index represents the annual composite radiative forcing by these gases relative to 1990, the Kyoto Protocol baseline year. The AGGI does not include indirect radiative effects (e.g., influences on ozone and water vapor). Based on the 2013 global mole fractions of LLGHGs and 15 minor gases there has been an additional 2.92 W [m.sup.-2] of direct radiative forcing since the pre-industrial era. The 2013 AGGI (Fig. 2.34) of 1.34 (2.92 W [m.sup.-2] / 2.18 W [m.sup.-2]) depicts a 34% increase since 1990 in the radiative forcing by the gases included in the AGGI.
2) OZONE-DEPLETING GASES--B. Hall, S. A. Montzka, G. Dutton, and J. W. Elkins
In addition to direct radiative forcing, long-lived gases containing chlorine and bromine also influence radiative forcing indirectly through destruction of stratospheric ozone. The atmospheric burdens of many of the most potent ozone-depleting gases have been declining in response to production and consumption restrictions imposed by the Montreal Protocol on Substances that Deplete the Ozone Layer and its Amendments (Figs. 2.32d, 2.33). Surface mole fractions of methyl chloroform (C[H.sub.3]C[Cl.sub.3]), which has a relatively short lifetime of five years, have declined 95% from peak values in the early 1990s (Fig. 2.33). Gases with longer lifetimes (Table 2.7) are declining more slowly.
Equivalent effective stratospheric chlorine (EESC) provides an estimate of the loading of ozone-reactive halogen in the stratosphere (and, therefore, the poten
tial to deplete stratospheric ozone). EESC is derived from surface measurements of ozone-depleting gases and weighting factors that include surface to stratosphere transport times, mixing during transit, photolytic reactivity, and bromine's enhanced efficiency in destroying ozone relative to chlorine (Fig. 2.35a; Schauffler et al. 2003; Newman et al. 2007; Montzka et al. 2011). Progress towards decreasing the stratospheric halogen load back to its 1980 level, a benchmark often used to assess ozone layer recovery, is evaluated by the NOAA Ozone-Depleting Gas Index (ODGI; Hofmann and Montzka 2009). The ODGI relates EESC in a given year to the peak and 1980 EESC values (Fig. 2.35b).
The EESC is calculated for two representative stratospheric regions (polar and middle latitudes) that differ in transit times and reactive halogen liberation. On average, it takes an air parcel about 3 and 5.5 years from its time of stratospheric entry (mainly in the tropics) to reach the ozone layer above the midlatitudes and poles, respectively. EESC values over the poles are significantly greater than over the midlatitudes because more halogen is liberated in transit to the polar region. At the beginning of 2013, EESC (ODGI) values were -3880 ppt (84) and -1650 ppt (62) over Antarctica and the midlatitudes of both hemispheres, respectively (Fig. 2.35a). The ODGI for Antarctica depicts a 16% decline in EESC towards the 1980 benchmark from its peak in 2001-02 through the beginning of 2013 (Fig. 2.35b). There was a 38% decline in ODGI over the midlatitudes from 1997 to the beginning of 2013. Both regions show similar decreases in EESC from peak values, but the relative ODGI scale indicates greater progress towards the 1980 benchmark for the midlatitude stratosphere because of the smaller difference between its peak and 1980 EESC values.
3) AEROSOLS--A. Benedetti, L. T. Jones, J. W. Kaiser, J.-J. Morcrette, and S. Remy
Atmospheric aerosols were prominently discussed in the recent Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (Boucher et al. 2013). While there is general agreement between the various estimates of the net impact of aerosols on the reduction of solar radiation due to reflection, there are still large uncertainties in the role of absorbing aerosols such as black carbon that have a net warming impact at the surface. It is noteworthy that atmospheric composition reanalyses are now appearing side-by-side with the pure modeling approach; these are often based on an ensemble of models (e.g., AEROCOM, http://aerocom.met.no). For example in AR5, one of the aerosol forcing estimates is based on the Monitoring Atmospheric Composition and Climate (MACC) reanalysis of satellite observations (Bellouin et al. 2013).
State-of-the-art aerosol reanalyses optimally incorporate satellite observations into atmospheric models that couple aerosol processes with the meteorology. Estimates of total aerosol optical depth (AOD) are well constrained by the satellite observations, while the speciated aerosol information is derived from the model. Limitations of current aerosol reanalyses include short time series, unknown biases in emission and removal processes, uncertainties in the optical properties of aerosols, and a limited availability of constraining satellite observations. However, such reanalyses represent a useful tool for assessing the current state of these important atmospheric constituents.
The MACC/MACC-II data assimilation system was used to produce a reanalysis of atmospheric composition, including aerosols, for the years 2003-12 (Morcrette et al. 2011; Inness et al. 2013). The aerosol model provides concentrations and optical depths for five species: desert dust, sea salt, organic matter, black carbon, and sulfate. All relevant physical processes such as emissions, wet/dry deposition, and sedimentation are included. The spatial resolution of this dataset is ~80 km.
The aerosol model underwent further development after the completion of this 10-year reanalysis with the aim of increasing its forecasting skill. Modifications to the dust parameterization in the new model version lead to larger amounts of mineral dust while changes in the meteorological model produce weaker surface winds over the remote oceans and a decrease in sea-salt aerosols. The modified model was used to produce an analysis for 2013 at a resolution of ~40 km that, like the 10-year reanalysis, used 4DVAR assimilation of AOD observations at 550 nm from the MODIS sensors, including a global adaptive bias correction. Biomass burning emissions for 2013 were provided by the Global Fire Assimilation System (GFAS) inventory (Kaiser et al. 2012) that estimates emissions from MODIS observations of fire radiative power. The 10-year reanalysis employs GFAS-based fire emissions for 2009-12 and the GFED3 inventory for 2003-08 (van der Werf et al. 2010).
Aerosol anomalies are determined by subtracting the multiyear average from the annual mean for a given year; however, due to the large changes in desert dust and sea salt in the recent model version, only the anomalies for carbonaceous aerosols are presented here. These are defined as the sum of anthropogenic organic matter, black carbon, and biomass burning emissions. Results are presented with a focus on the strongest seasonal anomalies in Northern Hemisphere summer (June-August, JJA) and fall (September-November, SON) 2013.
Global maps of the time-averaged total AOD and carbonaceous AOD from the MACC-II reanalysis are shown in Fig. 2.36 for the period 2003-12. Note the widespread pollution over most of Asia, particularly China and India, the large values over the Arabian Peninsula where the effects of dust and anthropogenic aerosols are compounded, the large biomass burning region in Central Africa, and the seasonal biomass burning signal in South America. The anomaly map for carbonaceous AOD for JJA 2013 is displayed in Fig. 2.37a. A striking feature of the boreal summer 2013 is smoke from the extensive burning in Canada being carried into the Atlantic Ocean towards Europe by a strong jet stream. At the same time more localized but intense fires were burning in Colorado. Figure 2.37a also shows the active burning season in Siberia and Central Africa. Figure 2.37b presents a similar plot for SON 2013. Noteworthy features include the persistent positive anomaly over Africa and the negative anomaly in biomass burning over South America, possibly connected to the decreasing trend in deforestation. The anomalies identified in the biomass-burning aerosol fields are consistent with the 2013 carbon monoxide and fire anomalies reported in sections 2g7 and 2h4 respectively.
4) STRATOSPHERIC OZONE--M. Weber, W. Steinbrecht, R. van der A, M. Coldewey-Egbers, V. E. Fioletov, S. M. Frith, C. S. Long, D. Loyola, and j. D. Wild
Annual mean total ozone values for 2013 were similar to the 1998-2008 decadal average in the midlatitudes and were above average in the tropics and high latitudes of both hemispheres (Plate 2.1p). The positive 2013 anomalies for both polar regions were greater than 25 Dobson Units (DU). Significant negative anomalies (below -10 DU) observed in the North Pacific near the Aleutian Islands are attributed to warmth associated with persistent ridging in the lower stratosphere over this region. The band of positive ozone anomalies surrounding the equator is a typical dynamical pattern caused by the westerly phase of the quasi-biennial oscillation (QBO).
The annual mean anomalies at middle to polar latitudes are largely determined by the ozone amounts during the winter and spring seasons. These depend on stratospheric meteorological conditions that exhibit strong interannual variability (Steinbrecht et al. 2011; Weber et al. 2011). Positive anomalies in the Southern Hemisphere polar region (see Plate 2.1p) are related to the rather weak and small ozone hole in winter/spring 2013 (see section 6g). Spring total ozone values for 2013 in the Arctic (March) and Antarctic (October) were close to and greater than their decadal averages, respectively (Fig. 2.38).
In Fig. 2.39, time series of total ozone since 1970 from different data sources are shown for several zonal bands: global (60[degrees]S-60[degrees]N), midlatitudes (35[degrees]-60[degrees]) in both hemispheres, and the tropics (20[degrees]S-20[degrees]N). The global average for 2013 is at the high end of the range of values observed since 2000, as might be expected from the 2013 maximum in solar cycle 24. As in the tropics, a quasi-biennial variation is also evident in the extratropics.
The large decline in global ozone between 1980 and the early 1990s was followed by a rapid increase and subsequent leveling off of Northern Hemisphere ozone values. In the Southern Hemisphere, total ozone shows no significant change since the early 1990s except for interannual variability. The substantial minima in Northern Hemisphere and global ozone in the early 1990s (Fig. 2.39) are related to enhanced levels of volcanic aerosols from Mount Pinatubo. The major eruption in 1991 provided additional surfaces for heterogeneous chemical reactions that remove ozone (WMO 1999). Total ozone values in the Southern Hemisphere did not show a similar minimum following Pinatubo because of additional long-wave radiative heating by aerosols, enhanced planetary wave activity (strengthening the Brewer Dobson circulation), and stronger QBO-induced downwelling in the Southern Hemisphere extratropics. These ozone-enriching mechanisms effectively compensated for the post-Pinatubo aerosol-related ozone depletion in the Southern Hemisphere (Schnadt Poberaj et al. 2011; Aquila et al. 2013).
Despite the considerable year-to-year variability, the leveling off of total ozone values since the mid1990s attests to the success of the Montreal Protocol and its Amendments in phasing out ozone-depleting substances (see section 2g2; e.g., Kiesewetter et al. 2010; Mader et al. 2010; Steinbrecht et al. 2011; Chehade et al. 2013; Frossard et al. 2013; Kuttipurath et al. 2013; Nair et al. 2013). Apart from variability in tropical ozone related to the 11-year solar cycle, QBO, and ENSO, there has been no statistically significant change in the tropics since the early 1990s (Chehade et al. 2013).
Upper stratospheric ozone has the largest sensitivity to changes in the stratospheric halogen loading due to the smaller influence of dynamical factors in that region. It is therefore easier to detect evidence of anthropogenic ozone recovery in the upper stratosphere (Newchurch et al. 2003) than in the lower stratosphere where most of the total ozone column resides. Ozone in the upper stratosphere (35-45 km) measured by ground-based and satellite instruments is shown in Fig. 2.40. Ozone decreased substantially by 10%-15% from the early 1980s until the mid-1990s (Steinbrecht et al. 2009; Jones et al. 2009). Since 2000, the various upper stratospheric records show signs of an increase of up to 5% through 2013 at most stations except Mauna Loa, Hawaii, where ozone has remained more or less constant since the mid-1990s (Fig. 2.40). Upper stratospheric ozone is influenced by changes in the atmospheric burdens of ODS and greenhouse gases (Fleming et al. 2011; Gillett et al. 2011). The observed long-term behavior of ozone in the upper stratosphere is consistent with the earlier increase and subsequent slow decline of the stratospheric halogen loading (Fig. 2.35).
5) STRATOSPHERIC WATER VAPOR--D. F. Hurst, S. M. Davis, and K. H. Rosenlof
Anomalies in tropical lower stratospheric water vapor were strongly negative (dry) at the start of 2013. Observations by the Aura Microwave Limb Sounder (MLS) during January 2013 depict tropical anomalies as large as -1.0 ppmv (-30%) at 82 hPa (Fig. 2.41a). By July, the dry tropical anomalies had weakened but had also spread out globally in the lower stratosphere (Fig. 2.41b). In general, the dry anomalies in July 2013 were stronger and more globally pervasive than in July 2012 (see figure 2.37 in Hurst and Rosenlof 2013). However, in January 2013 there were contrasting positive (wet) anomalies over the high latitudes of each hemisphere. The Arctic anomalies may be related to the strong sudden stratospheric warming event in January 2013 (see section 2b3) accompanied by enhanced downwelling of older, wetter air into the lower stratosphere. Wet anomalies over the high southern latitudes are attributed to relatively weak dehydration within the smaller and warmer Antarctic vortex in 2012 that had split into two parts by early November (Long and Christy 2013; Newman et al. 2013).
The seasonal variability of water vapor abundance in the lower tropical stratosphere is predominantly controlled by the annual cycle of cold-point tropopause temperatures in the tropical tropopause layer (TTL). The resulting cycle in water vapor is visible in the tropical MLS data at 100 hPa (Fig. 2.42a). The erosion of this annual cycle in water vapor is evident as the Brewer-Dobson circulation carries tropical air masses upwards. Interannual variations in tropical lower stratospheric water vapor are more overt when viewed as anomalies. In October 2012 the weak dry anomaly at 100 hPa intensified, remained strong until early 2013, then weakened (Fig. 2.42b). This behavior is consistent with cold-point tropopause temperatures (Fig. 2.43c, blue curve) reaching a six-year minimum in late 2012 due to the downward propagation of an easterly (cold) QBO phase into the TTL and then increasing throughout 2013 as the QBO phase became westerly.
The Aura MLS has now amassed near-global stratospheric water vapor measurements since August 2004. These data can be combined with measurement records from balloon-borne frost point hygrometers to evaluate trends. Figure 2.43 presents time series of zonally- and monthly-averaged MLS retrievals and data from quasi-monthly soundings of NO A A frost point hygrometers (FPH) at Boulder, Colorado; Hilo, Hawaii; and Lauder, New Zealand, and of cryogenic frost point hygrometers (CFH) at San Jose, Costa Rica. Agreement between the FPH and MLS at 82 hPa ranges from [+ or -]0.3% at Lauder to [+ or -]6% at Hilo. Deeper in the stratosphere (68-26 hPa) the long-term FPHMLS agreement is better than 1% (Hurst et al. 2014).
The 2013 anomalies at 82 hPa over the Hilo and San Jose tropical sites (Figs. 2.43b,c) depict increases of 0.4-0.8 ppmv (15%-25%) after two-year decreases of 1.2-1.4 ppmv (40%-45%). The expectation that the tropical increase would propagate to the middle latitudes within several months was not entirely met (Fig. 2.43); the 2013 data instead show only a weak increase of 0.2 ppmv at Boulder and a decrease of 0.3 ppmv at Lauder (Figs. 2.43a,d). Lower stratospheric water vapor over these two sites was evidently influenced by more than just transport from the tropics. The timing of the 2013 increase at Boulder implicates a tropical source but the substantially weakened tropical signal suggests other influences. For Lauder at 82 hPa, dry remnants of the 2013 Antarctic vortex may have reduced water vapor mixing ratios more than they were enhanced by the tropical increase.
6) TROPOSPHERIC OZONE--0. R. Cooper and J. Ziemke
The State of the Climate in 2012 report marked the first appearance of tropospheric ozone in this annually recurring series (Cooper and Ziemke 2013), with a summary of 1990-2010 surface and free-tropospheric ozone trends around the globe based on in situ observations reported in the peerreviewed literature. A similar summary for 2013 is not possible due to the absence of any systematic procedure for routinely updating ozone trends based on in situ observations at the surface and in the free troposphere. However, procedures are in place for timely updates to tropospheric column ozone (TCO) derived from the Ozone Monitoring Instrument (OMI) and Microwave Limb Sounder (MLS) remote sensing instruments onboard NASA's polar orbiting Aura satellite (Ziemke et al. 2006, 2011). Therefore, this assessment of tropospheric ozone in 2013 relies on the OMI/MLS TCO product, spanning 2005-13 (Figs. 2.44, 2.45).
The average 2013 tropospheric ozone burden (60[degrees]S-60[degrees]N) from OMI/MLS was 275 Tg, 2.5% above the 2005-12 mean of 268 Tg (Fig. 2.45; Plate 2.1o). Relative enhancements during 2013 in the Northern Hemisphere (NH) and Southern Hemisphere (SH) were similar at 2.8% and 2.2%, respectively, but the seasons and latitudes at which the enhancements occurred differ hemispherically (Fig. 2.45). The strongest positive anomalies of 6.6%, 5.0%, and 5.1% in the NH were in the tropical latitudes during winter (DJF), spring (MAM), and summer (JJA), respectively. In the SH, the strongest enhancements occurred in the extratropics with anomalies of 5.8%, 4.0%, and 4.1% in summer (DJF), fall (MAM), and winter (JJA), respectively. There was a negative anomaly of-4.1% in the SH tropics during spring (SON). Simulations of tropospheric ozone by chemical transport models indicate that the burden fluctuates interannually and seasonally due to variability in: (1) ozone transport from the stratosphere (Ordonez et al. 2007; Voulgarakis et al. 2011; Hess and Zbinden 2013); (2) photochemical processes modulated by large-scale meteorology driven by ENSO (Doherty et al. 2006; Koumoutsaris et al. 2008; Voulgarakis et al. 2010); and (3) variability in biomass burning emissions (Leung et al. 2007; Sauvage et al. 2007). Modelling studies have not yet been conducted to determine the causes of the 2013 ozone anomalies.
From October 2004 through December 2013, the global tropospheric ozone burden increased significantly at a linear rate of 1.9 [+ or -] 0.9 Tg [yr.sup.1] (p<0.01), with growth rates of 0.7 [+ or -] 0.8 Tg [yr.sup.1] (p=0.08) and 1.2 [+ or -] 1.0 Tg [yr.sup.1] (p=0.01) in the NH and SH, respectively (Fig. 2.45). Since tropospheric ozone abundance is influenced by the ENSO cycle, the relatively short OMI/MLS time series precludes the conclusion that the linear increase in the tropospheric ozone burden is part of a long-term trend. Several more years of data are required to confidently detect a trend beyond the noise associated with meteorological cycles. Chemistry-climate models would then be required to attribute any observed trends to changes in emissions (anthropogenic or natural; Young et al. 2013; Parrish et al. 2014; Cooper et al. 2014, manuscript submitted to Elementa), transport patterns (Lin et al. 2014), meteorology (Voulgarakis et al. 2010), or influence from the stratosphere (Hess and Zbinden 2013).
7) CARBON MONOXIDE--J. Flemming and A. Inness
Though carbon monoxide (CO) is not a direct climate forcing agent it influences the abundance of greenhouse gases like methane ([CH.sub.4]) through hydroxyl radical (OH) chemistry and plays an important role in the production of tropospheric ozone (Hartmann et al. 2013). CO is emitted to the atmosphere during incomplete combustion of fossil fuels and biomass and is produced in situ by the oxidation of CH4 and other organic trace gases. Combustion and in situ sources typically produce similar amounts of CO each year. CO has a lifetime of 1-2 months and is therefore a good indicator of the long-range transport of pollutants.
The Monitoring of Atmospheric Composition and Climate (MACC) data assimilation system provides analyses and forecasts of atmospheric composition (Inness et al. 2013). A reanalysis of atmospheric composition for 2003-12 and a near-real time analysis for 2013 assimilated total column CO retrievals between 65[degrees]N and 65[degrees]S from MOPITT (Deeter et al. 2010; Deeter 2011) from 2003 onwards (Version 4 during 2003-12, Version 5 in 2013), and between 70[degrees]N and 70[degrees]S from IASI (George et al. 2009; Clerbaux et al. 2009) since April 2008. The satellite observations were assimilated in the ECMWF's Integrated Forecasting System, which was coupled to the chemical transport model MOZART-3 (Kinnison et al. 2007) as described in Flemming et al. (2009). The anthropogenic emissions of the assimilating model were taken from the MACCity inventory (Granier et al. 2011) that accounts for projected trends in the emissions. Biomass burning emissions from the GFED (v3.0) inventory (van der Werf et al. 2010) were used for the years 2003-08. Since 2009 daily biomass burning emissions from MACC's GFAS, Version 1.0 (Kaiser et al. 2012) have been employed. The MACC CO dataset provides the global three-dimensional distribution of CO and is used here to assess CO total column anomalies for 2013.
The climatological distribution of total column CO for 2003-13 is presented as the medians of monthly averages during this period (Fig. 2.46). In general there is a hemispheric gradient in total column values, with 2 x [10.sup.18] molecules [cm.sup.-2] in the Northern Hemisphere and 1 x [10.sup.18] molecules [cm.sup.-2] in the Southern Hemisphere. Regional maxima are located over central Africa due to intensive biomass burning and over Southeast Asia and southern Asia because of strong anthropogenic emissions. Outflow from these regions increases CO column values over adjacent regions in the eastern Atlantic and the western Pacific. In the Southern Hemisphere, away from the main emissions sources, the typical background CO mixing ratios of 40-60 ppb vary little with height in the troposphere. In the Northern Hemisphere and the biomass burning regions of the tropics, typical mixing ratios range from ~60 ppb in the upper troposphere to about 150 ppb near the surface. In regions of high emissions the CO mixing ratios can be more than 10 times these typical values.
Without further calibration, changes to the assimilated satellite retrievals and the relatively short period covered by the MACC CO reanalysis make it insufficiently consistent for the investigation of long-term trends. During 2003-13, the MACC reanalysis shows CO trends of-0.7% and -0.9% [yr.sup.-1] for the globe and Northern Hemisphere, respectively, if estimated by linear fits. Trends of -1% [yr.sup.-1] were also found for the globe and Northern Hemisphere in a study of CO measurements by different satellite-based instruments during the last decade (Worden et al. 2013).
To investigate the spatial distribution of the anomalies in 2013, a latitude-dependent bias correction was applied to the climatological distribution to remove long-term trends. No regional absolute anomalies >10% were found for the 2013 average CO columns. On seasonal time scales, a positive CO anomaly >0.3 x [10.sup.18] molecules cm-2 (>20%) occurred over Northern Siberia in June-August (JJA) due to large fires in July (Fig. 2.47a). During this period, there was also a positive anomaly over tropical Africa due to seasonal biomass burning (see section 2h4). In 2013, the fire season in tropical South America was less intense than the decadal average, leading to negative anomalies of more than -0.2 x [10.sup.18] molecules [cm.sup.-2] (-10%) in both JJA (Fig. 2.47a) and September-November (SON) as well as less CO outflow over the Southern Atlantic Ocean in SON (Fig. 2.47b).
h. Land surface properties
I) FOREST BIOMASS--S. Quegan, P. Ciais, and M. Santoro
While inventories and in situ measurements form the basis of much of our knowledge about the worldwide values and distribution of biomass, there are enormous data gaps, especially in the tropical belt, and also regarding the current status of the vast Eurasian forests. Although mechanisms such as the UNFCCC Reducing Emissions from Deforestation and Forest Degradation initiative (REDD+) have spurred efforts to establish national forest inventory systems in several tropical countries, the recent production of regional to continental scale biomass datasets has relied on satellite data analysis and coordination of in situ measurements. There has also been considerable effort devoted to testing satellite-derived pan-tropical biomass maps using both airborne lidar and in situ data.
Methods to derive biomass in boreal and temperate forests from long time series of C-band Envisat satellite radar data (Santoro et al. 2011) have been applied to estimate the carbon stock of Northern Hemisphere forests north of 30[degrees]N as of 2010 (Thurner et al. 2014; Fig. 2.48). Although available at 0.01[degrees] resolution (http://www.biomasar.org), Santoro et al. (2013) demonstrate that accuracy at this scale is comparatively poor, and spatial averaging provides more reliable results: at 0.5[degrees] spacing, estimated growing stock volume has a relative accuracy of 20%-30% when tested against inventory data. The associated global estimates of carbon stored in boreal, temperate mixed and broadleaf, and temperate coniferous forests are 40.7,24.5, and 14.5 Pg C respectively, with estimated accuracies of around 33%-40% (Thurner et al. 2014).
Continued assessment of the pan-tropical biomass maps of Saatchi et al. (2011) and Baccini et al. (2012) has revealed significant regional differences between them, although when aggregated to country or biome scale these disagreements tend to decrease (Mitchard et al. 2013). Of current concern is the lack of satellites in orbit capable of providing information on forest biomass. The biomass maps produced by Santoro et al. (2013) and Thurner et al. (2014) are derived from a long time series of C-band radar data produced by the Envisat ASAR instrument, which failed in April 2012. Estimates of the biomass in lower biomass tropical woodlands (e.g., Mitchard et al. 2009) relied on the Japanese Space Agency (JAXA) ALOS-PALSAR L-band radar, which failed in May 2011. Both the Baccini et al. (2012) and Saatchi et al. (2011) tropical biomass maps are based on height measurements from Icesat, which failed in 2009. The situation should ease with the launch of the European Sentinel-IA and IB C-band satellites in spring 2014 and 2015, respectively, and the JAXA ALOS-2 L-band radar in spring 2014. An important development was the selection by the European Space Agency in May 2013 of the BIOMASS mission, a P-band radar dedicated to global forest biomass measurements (European Space Agency 2012), but this will not launch before 2020.
Local studies in the boreal zone have used allometric relations between biomass and forest height measured by the TanDEM-X satellite to derive biomass with root mean square accuracy between 16% and 20% at stand level (Askne et al. 2013; Solberg et al. 2013). This technique requires an accurate digital elevation model, which is not available at global scale, in particular for dense tropical forests; hence, it may not be possible to infer biomass worldwide from the 90-m global dataset expected to be soon available from TanDEM-X.
Collation of above-ground biomass data from 260 sample plots across 12 countries in intact closed canopy tropical forests in west, central, and east Africa sampled between 1978 and 2012 (Lewis et al. 2013) closes a major gap in current knowledge, and reveals that these forests are characterized by high biomass values (mean value 396 Mg ha1) concentrated in fewer trees per unit area compared to either Amazonia or Borneo. Regression indicates that both climate and soil properties affect the biomass of the region; in particular, high temperatures and high rainfall in the wettest months are both associated with lower biomass.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||p. Sxiv-S45|
|Publication:||Bulletin of the American Meteorological Society|
|Date:||Jul 1, 2014|
|Previous Article:||Professional directory: space reserved for certified consulting meteorologists.|
|Next Article:||State of the climate in 2013.|