Longwave radiation at the earth's surface in Estonia/Aluspinna ja atmosfaari pikalaineline kiirgus Eestis.
The net radiation at the earth's surface is of dominant importance in studies of climate changes. Often the short- and longwave contributions to the radiation budget have been studied separately. Solar radiation has been continuously measured in many geographical sites during long periods. However, similar studies of long-wave radiation started mostly after the Baseline Surface Radiation Network (BSRN) was established as a project of the World Climate Research Programme (WCRP) in 1992. In Estonia continuous solar radiation measurements began in 1950. The characteristic features of shortwave components of net radiation in Estonia, found based on long time series, are described in the handbook by Russak and Kallis (2003). Systematic recordings of longwave radiation were started here only about ten years ago, and due to the shortness of these time series, no results of corresponding analysis have been published hitherto.
During the last decades, interest in the role of longwave radiation has increased considerably. This can be due to the role of atmospheric longwave radiation in enhancing the greenhouse effect, resulting in global warming. As thermal radiation depends on several meteorological parameters, e.g. on temperature and humidity and their profiles, cloudiness, atmospheric aerosol content, etc., infrared radiation varies both seasonally and geographically. At present, the amount of collected data on longwave radiation at several sites is sufficient to highlight its characteristic features.
In the present study the hourly totals (MJ [m.sup.-2]) of up-and downwelling longwave radiation recorded at Tartu-Toravere meteorological station of the Estonian Environment Agency (58[degrees]16'N, 26[degrees]28'E, h = 70 m a.s.l.) are used to study the features of the longwave components of the near-surface radiation budget in Estonia. Silicon-domed Eppley PIR pyrgeometers, installed at 2 m above a grassy surface, were used as receivers of longwave radiation. The instrument for recording downwelling atmospheric radiation [L.sub.1][down arrow] was shaded from direct solar rays by a tracking shading disc. Initially the receivers were calibrated in the World Radiation Centre in Davos, Switzerland. Continuous recording of downwelling atmospheric radiation [L.sub.1] [down arrow] began in 2003, that of upwelling fluxes ([L.sub.1] [up arrow]) in July 2006. The meteorological data necessary for analysis, i.e. the hourly mean bare soil temperature t ([degrees]C) and water vapour pressure e (hPa), were obtained using a Vaisala automatic weather station AMS520. The genera and amount of high, middle, and low clouds were visually determined in tenths at 1-hour intervals during daylight hours and at 3-hour intervals at night. Tartu-Toravere station is located in a rural area with no outdoor lighting. In our case, we used extreme conditions (cloudless and 10 tenths of low clouds), which are at night reliably determined by the visibility of stars. At Toravere an observer-staffed station operates with 24-hour maintenance of instruments.
3. UPWELLING LONGWAVE RADIATION
The upwelling longwave radiation [L.sub.1] [up arrow] depends mainly on surface temperature and slightly on surface properties. In the range of surface temperatures t observed at Toravere in 2007-2012 (from -27 to +32[degrees]C) the 2045 hourly totals of upwelling longwave radiation were credibly described by a linear relationship (Eq. 1), (Fig. 1):
[L.sub.1] [up arrow] = 0.017 t + 1.146, [R.sup.2] = 0.98. (1)
The significance levelp < 0.01.
According to the Stefan-Boltzmann law, the upwelling longwave radiation is proportional to the fourth power of the absolute temperature [T.sup.4] of ground surface. The linear dependence in our results is due to the circumstance that the Stefan-Boltzmann law holds for a full range of temperatures, but for narrow temperature ranges it reduces well to a linear dependence. To compare the consistence of measured and calculated [L.sup.1] [up arrow] hourly totals separately, we analysed the observational data recorded at hours with snow cover (depth of snow at least 3 cm). The radiation emissivity [epsilon] of snow depends on its structure and varies from 0.8 to 0.9 (e.g. www.monarchserver.com,www.omega.com). Our calculations for 534 hourly totals were carried out with the Stefan-Boltzmann formula, adopting [epsilon] = 0.85. We plotted the measured [L.sub.1] [up arrow] values against those calculated by the Stefan-Boltzmann law and found a good linear relationship ([R.sup.2] = 0.96). However, the measured totals of [L.sub.1] [up arrow] systematically exceeded the calculated values (on average by 18%).
In case of grassy surface, this kind of comparison was impossible. Problems emerged how to estimate the temperatures. The ground temperatures were measured on bare soil and, as known, the soil temperatures are typically higher than those of grass. Unfortunately, no reasonable relation between soil and grass temperatures was available to us.
In Estonia the annual totals of the longwave radiation emitted by the ground surface constituted on average 11 071 MJ [m.sup.-2] during the period 2007-2012. The proportions of different monthly totals in the annual sum vary on average from 6% (in February) to 10% (in July) (Table 1).
The inter-annual fluctuations of the [L.sub.1] [up arrow] monthly totals are small: in winter the coefficient of variation V = 6-9%, in summer 1-3%. The hourly totals pose a greater challenge (Fig. 2). The corresponding values of the coefficient of variation V = 12-14% in winter and 9-11% in summer. The greatest hourly total of [L.sub.1] [up arrow], 1.923 MJ [m.sup.-2], was recorded on 8 July 2011, the smallest, 0.609 MJ [m.sup.-2], on 5 February 2012.
The diurnal course of the [L.sub.1] [up arrow] hourly totals is rather marginal in winter months. The difference between their values in day- and night-time hours is evident only in the warm season, when incident solar radiation heats the ground. The difference between noon and midnight surface temperatures is practically lacking in winter, but in summer its diurnal amplitude is up to 15[degrees]C on average. The greatest average difference (about 20%) between the hourly sums of [L.sub.1] [up arrow] at noon and at midnight was observed in April (Fig. 3). This is evidently due to the diurnal course of ground temperature. In Estonia April is the main transitional time from cold to warm season conditions, when the ground is usually covered with old grass or snow, or both.
The shape of the frequency distribution over the set of the recorded hourly totals is asymmetrical. While all the values varied within the range from 0.61 to 1.92 MJ [m.sup.-2], about half of them were in the interval 1.13-1.37 MJ [m.sup.-2] (Fig. 4).
The distribution varies during the year, being almost symmetric in summer, but strongly skewed towards smaller values in the winter months. In winter a weak secondary maximum is noticeable, corresponding to frosty days (Fig. 5).
Besides longwave radiation, the earth's surface also loses radiation energy as reflected solar radiation [E.sub.r] [up arrow].
However, the role of the latter in the completely upwelling radiation is small. About 92% of the annual totals of the leaving radiation is due to [L.sub.1][up arrow] . Only in March, when the combined impact of increased solar radiation and high albedo of snow abruptly increases the reflected solar radiation, does the role of [L.sub.1] [up arrow] fall to 85% (Fig. 6).
4. DOWN WELLING LONGWAVE RADIATION
In comparison with upwelling longwave radiation, the downwelling atmospheric radiation [L.sub.1] [down arrow] is more variable due to its dependence on several factors (e.g. [L.sub.1] [up arrow] ; air temperature; the content of water vapour, C[O.sub.2], [O.sub.3], [CH.sub.4], [N.sub.2]O, CF, as well as some kinds of aerosols and also their vertical profiles). The annual total of [L.sub.1] [down arrow] averaged over the period 2003-2012 added up to 9808 MJ [m.sup.-2] at Toravere. Its inter-annual variability is small (the coefficient of variation V = 0.01) and the proportion of different monthly totals in annual totals rises from 6-7% in winter to 9-10% in summer (Table 2).
The diurnal course of the hourly totals of the atmospheric longwave radiation is small: over a 24-hour period, hourly totals change relatively just a bit. The difference between the totals in daytime and at night is practically lacking in winter. Even in summer, the totals of [L.sub.1] [down arrow] at noon exceed their values at midnight only by about 8% (Fig. 7). Its greatest hourly total, 1.584 MJ [m.sup.-2], was recorded on 15 July 2010, the smallest, 0.472 MJ [m.sup.-2], on 19 January 2006.
The frequency distribution of the hourly totals of [L.sub.1] [down arrow] is monomodal and negatively skewed (Fig. 8). Over the course of a year, the shape of the distribution varies basically as follows: in summer it is monomodal and in January-February bimodal, where the mode at smaller values corresponds to the cloudless frosty hours (Fig. 9).
The contribution of [L.sub.1] [down arrow] to the downward component of the radiation balance is greater than that of solar radiation (Fig. 10). Annual totals of [L.sub.1] [down arrow] exceed those of global radiation [E.sub.g] [down arrow] about three times. Over the course of a year this ratio, that is [L.sub.1] [down arrow]/[E.sub.g] [down arrow], varies largely.
In winter longwave radiation is crucial in the incoming radiation: from November to January its role is as high as 95-97%. From May to July the contributions of [L.sub.1] [down arrow] and [E.sub.g] [down arrow] become closer, and then the totals of atmospheric radiation comprise about 60% of the entire downwelling radiation. This type of great annual variability is mainly caused by the annual cycle of solar radiation. The latter in turn is highly dependent on the annual course of solar elevation and the duration of light time. In Estonia the solar elevation reaches about 55[degrees] and light time duration is 18 hours at noon on the summer solstice, on the winter solstice the corresponding values are only about 8[degrees] and 6 hours. In addition, the annual courses of the amount and genera of clouds play an important role (Fig. 11, Table 3). The optically thick low clouds, which intensively attenuate solar radiation simultaneously increasing atmospheric radiation, have the opposite impact on short- and longwave radiation.
5. FACTORS AFFECTING DOWNWELLING LONGWAVE RADIATION
Among the numerous factors affecting the fluxes of downwelling atmospheric radiation, water vapour is the dominant one. Although its emissivity is lower than, for example, that of C[O.sub.2] or [CH.sub.4], water vapour plays an essential role in the generation of counter-radiation fluxes owing to its higher abundance in the atmosphere. Due to the temporal and spatial variability of the atmospheric water content, the incoming longwave radiation can differ geographically. The atmospheric radiation depends not only on the amount of water vapour, but also on its vertical profile and temperature. No suitable data on the corresponding gradients were at our disposal. Therefore we decided to use here the relationship between near-surface water vapour pressure e (hPa) and hourly totals of downwelling longwave radiation [L.sub.1] [down arrow] (MJ [m.sup.-2]). The choice was made taking into account that the near-ground humidity parameters depend linearly on the total amount of water vapour in the vertical air column (e.g. Okulov, 2003; Ruckstuhl et al., 2007; Kannel et al., 2012) and the lowermost air layers are dominant for the vertical gradient of water vapour.
From analysis of the corresponding data for 20 355 hours from the period 2006-2012 it followed that the dependence between e and [L.sub.1] [down arrow] was best described by a power function (Eq. 2), (Fig. 12):
[L.sub.1] [down arrow] = 0.704 [e.sup.0.23], where [R.sup.2] = 0.73. (2)
Our result fits well the relationships between specific humidity and incoming longwave radiation obtained from the analysis of measurement data from BSRN as well as some other actinometrical stations (Ruckstuhl et al., 2007; Stanhill, 2011; Rosa and Stanhill, 2014). A slightly different result from our Eq. (2) was obtained for Toravere observations by Rosa and Standhill (2014) (exponent b = 0.22). It is interesting to note that we found a good accordance between the results by Stanhill (2011) for Valentia (Ireland) and our results for Toravere. It is likely that this can partially be explained by similar air humidity and cloudiness conditions at these sites.
In Estonia the power relationship between e and [L.sub.1] [down arrow] is stronger in winter than in summer ([R.sup.2] = 0.76-0.78 in January-February and [R.sup.2] = 0.41-0.42 in June-July). This is most likely caused by the differences in the mean vertical profiles of water vapour and temperature in cold and warm seasons. In the latter case, intensive air convection and turbulence carry humidity from the near-surface layers upwards, thus decreasing the relative role of the lowermost layers in the process of [L.sub.1] [down arrow] formation. In addition, the variations in cloudiness conditions in the course of a year affect this dependence.
In order to estimate the contribution of clouds to the fitted power function, we analysed the data of the hours with 10 tenths of low clouds (7950 hours) separately from those of the hours for clear sky (2328 hours). In both cases the power relationship holds, but a difference was found in their parameters (Fig. 13). The corresponding approximations are
[L.sub.1] [down arrow] = 0.796 [e.sup.0.20], [R.sup.2] = 0.93, overcast with low clouds; (3)
[L.sub.1] [down arrow] = 0.601 [e.sup.0.25], [R.sup.2] = 0.91, clear sky. (4)
Within the limits of water vapour pressure in Estonia (0.5-25 hPa), 10 tenths of low clouds enlarge the counter-radiation on average by 12-37%. This effect is greater at smaller values of e, which are characteristic for the winter months. During cloudless hours about 91% of the variations in atmospheric downward radiation in Estonia can be attributed to the variations in the near-ground water vapour pressure. It is remarkable that also this effect is greater in winter months. In summer [R.sup.2] remains in the range of 0.5-0.6 (Table 4).
The aim of the present study is to describe the features of longwave radiation in Estonia and their dependence on the ground temperature, water vapour pressure, and cloudiness. Here we present the main conclusions.
In the annual totals of the incoming radiation, the atmospheric downwelling longwave radiation exceeds solar radiation about three times. This ratio has an essential seasonal course, which depends mainly on the annual solar cycle (solar elevation and the duration of light time) and on cloudiness conditions. In the annual totals of upwelling radiation, the proportion of longwave radiation is even greater than it is for downwelling radiation, constituting on average 92%. In comparison with solar radiation, the totals of longwave radiation are more stable.
In the limited range of ground temperatures t in Estonia during 2007-2012 (from -27 to +32[degrees]C) the dependence of [L.sub.1] [up arrow] on t is linear ([R.sup.2] = 0.98). Comparison of the measured [L.sub.1] [up arrow] values and those calculated by the Stefan-Boltzmann law for snow (emissivity [epsilon] = 0.85) showed a good linear relationship ([R.sup.2] = 0.96). However, the measured totals of [L.sub.1] [up arrow] were systematically higher (on average by 18%).
The downwelling atmospheric radiation [L.sub.1] [down arrow] depends mainly on the atmospheric water vapour distribution. Between the hourly totals of the atmospheric radiation and near-surface water vapour pressure e, a power function fitted. The parameters of the relationship vary considerably, depending on the cloudiness conditions. Within the limits of e observed in Estonia (0.5-25 hPa), 10 tenths of low clouds increase the downwelling atmospheric radiation on average by 12-37%. This effect is greater at smaller values of humidity, typical in Estonia during the winter months. In cloudless hours, about 91% of the variation in atmospheric radiation is connected to the variations in the near-ground water vapour pressure. It is characteristic that the dependence of [L.sub.1] [down arrow] on e is stronger in winter months. This is most likely caused by differences in the mean vertical profiles of moisture and temperature in cold and warm seasons.
As our radiation time series is short, we could not find any statistically significant trends either for [L.sub.1] [up arrow] or for [L.sub.1] [down arrow]. Our results, obtained based on this short time series and considering the small temporal variability of longwave radiation, fix the current main characteristic features of longwave radiation in local conditions. Tartu-Toravere meteorological station is one of the few places in North Europe where components of the longwave radiation balance have been measured. As the weather in the Baltic Sea region is determined mainly by the cyclones moving eastwards from the Atlantic Ocean, the conditions of temperature, humidity, cloudiness, and other meteorological parameters that determine longwave radiation are quite similar all over Estonia. We even surmise that the description of near-surface longwave radiation presented in this paper can approximately be characteristic not only for Estonia, but also for the whole territory around the Baltic Sea.
We are grateful to the European Regional Development Fund for supporting the investigation within the project 'Estonian Radiation Climate'. We thank the Estonian Environment Agency for making available the radiation and meteorological database of Tartu-Toravere meteorological station.
Kannel, M., Ohvril, H., and Okulov, O. 2012. A shortcut from broadband to spectral aerosol optical depth. Proc. Estonian Acad. Sci., 61, 266-278.
Okulov, O. 2003. Variability of Atmospheric Transparency and Precipitable Water in Estonia During the Last Decades. PhD dissertation. Tartu University Press. Rosa, R. and Stanhill, G. 2014. Estimating long-wave radiation at the Earth's surface from measurements of specific humidity. Int. J. Climatol., 34, 1651-1656.
Ruckstuhl, C., Philipona, R., Morland, J., and Ohmura, A. 2007. Observed relationship between surface specific humidity, integrated water vapor, and longwave downward radiation at different altitudes. J. Geophys. Res.-Atmos., 112(D3), D03302.
Russak, V. and Kallis, A. 2003. Eesti kiirguskliima teatmik [Handbook of Estonian Solar Radiation Climate], Eesti Vabariigi Keskkonnaministeerium, Tallinn (in Estonian).
Stanhill, G. 2011. The role of water vapor and solar radiation in determining temperature changes and trends measured at Armagh, 1881-2000. J. Geophys. Res.Atmos., 116, D03105.
Viivi Russak (a) * and Ingrid Niklus (b)
(a) Tartu Observatory, 61602 Toravere, Tartumaa, Estonia (b) Estonian Environment Agency, Mustamae tee 33, 10616 Tallinn, Estonia
Received 24 January 2015, revised 20 March 2015, accepted 25 March 2015, available online 26 November 2015
* Corresponding author, firstname.lastname@example.org
Table 1. Mean monthly totals of upwelling longwave radiation [L.sub.l][up arrow] and their standard deviations (MJ [m.sup.-2]) at Toravere, 2007-2012 Jan Feb March Apr May June July Totals 786.3 688.9 837.1 902.2 1022.5 1043.6 1123.6 SD 53.8 64.9 33.8 12.8 12.0 24.8 34.0 Aug Sep Oct Nov Dec Totals 1085.6 975.2 938.2 851.9 816.5 SD 26.9 12.8 27.9 19.9 47.8 Table 2. Mean monthly totals of downwelling longwave radiation [L.sub.1][down arrow] and their standard deviations (MJ [m.sup.-2]) at Toravere, 2003-2012 Jan Feb March Apr May June July Totals 740.4 638.5 732.1 747.2 847.4 873.7 963.1 SD 40.5 54.6 49.1 17.3 25.8 14.8 32.3 Aug Sep Oct Nov Dec Totals 955.1 870.5 853.6 802.7 780.7 SD 19.4 14.3 32.5 24.7 35.8 Table 3. Mean hourly totals of downwelling longwave radiation [L.sub.1][down arrow] and the corresponding standard deviations (MJ [m.sup.-2]) in cloudless conditions (y = 0) and in overcast by low clouds (m = 10) conditions in Estonia, 2003-2012 Jan Feb March Apr May June [L.sub.1][down arrow] 0.655 0.675 0.823 0.907 1.012 1.113 (y = 0) SD 0.102 0.077 0.097 0.083 0.097 0.092 [L.sub.1][down arrow] 1.087 1.068 1.122 1.176 1.258 1.328 (m = 10) SD 0.083 0.087 0.075 0.056 0.075 0.061 July Aug Sep Oct Nov Dec [L.sub.1][down arrow] 1.225 1.176 1.042 0.924 0.860 0.808 (y = 0) SD 0.110 0.100 0.078 0.065 0.078 0.074 [L.sub.1][down arrow] 1.404 1.390 1.314 1.242 1.173 1.106 (m = 10) SD 0.054 0.061 0.066 0.072 0.069 0.088 Table 4. Parameters of the power relationship between hourly totals of downwelling longwave radiation [L.sub.1][down arrow] (MJ [m.sup.- 2]) and hourly mean water vapour pressure e (hPa) in different months in cloudless and overcast with low clouds conditions at Toravere, 2006-2012 Covered by 10 tenths of low clouds a b [R.sup.2] Hours Jan 0.798 0.20 0.87 1121 Feb 0.807 0.19 0.90 822 March 0.763 0.23 0.82 705 Apr 0.814 0.19 0.71 431 May 0.814 0.19 0.74 269 June 0.792 0.20 0.73 336 July 0.756 0.22 0.62 191 Aug 0.728 0.23 0.70 353 Sep 0.769 0.21 0.70 497 Oct 0.785 0.21 0.80 844 Nov 0.778 0.21 0.81 1148 Dec 0.800 0.20 0.88 1234 Cloudless a b [R.sup.2] Hours Jan 0.601 0.20 0.89 150 Feb 0.625 0.17 0.71 229 March 0.583 0.26 0.80 249 Apr 0.651 0.21 0.51 308 May 0.684 0.20 0.51 303 June 0.580 0.27 0.52 231 July 0.546 0.30 0.62 211 Aug 0.544 0.30 0.64 203 Sep 0.550 0.28 0.80 192 Oct 0.569 0.26 0.75 116 Nov 0.572 0.25 0.71 66 Dec 0.641 0.17 0.84 76
|Printer friendly Cite/link Email Feedback|
|Author:||Russak, Viivi; Niklus, Ingrid|
|Publication:||Proceedings of the Estonian Academy of Sciences|
|Date:||Dec 1, 2015|
|Previous Article:||Temperature extremes and detection of heat and cold waves at three sites in Estonia/Ekstreemsete temperatuuride esinemistoenaosused ning kuuma- ja...|
|Next Article:||Thermal annealing of sequentially deposited SnS thin films/SnS kilede keemiline sadestamine ZnS ja CdS alusele ning nende jargnev termiline...|