A perspective on bottom water temperature anomalies in Long Island Sound during the 1999 lobster mortality event.

ABSTRACT Analyses of time series data for bottom or near bottom temperatures for 50 stations distributed throughout Long Island Sound reveal distinctive features of the bottom water temperature history during the lobster mortality event of 1999. These include: temperatures that exceeded 23.5[degrees]C in shallow, well-mixed areas; markedly higher temperatures, in general, in those areas with water column depth <20 m; basin-averaged bottom temperatures that were the highest for the decade during the months of July and August; and a rapid increase in bottom temperatures in late August caused by the vertical mixing of warm surface waters during a strong wind event. Results indicate that anomalies in the local surface heat flux made an important contribution to bottom temperature anomalies.

KEY WORDS: temperature, climatology, heat flux, lobster, Homarus americanus

INTRODUCTION

The purpose of this article is to report results from analyses of physical water column properties in Long Island Sound (L1S) during the summer and fall of 1999 when a significant mortality of the American lobster (Homarus americanus H. Milne Edwards, 1837) took place. We examined water column temperature, salinity and dissolved oxygen, but we restrict the discussion here primarily to bottom temperature because we consider it the most directly relevant. Our objective is to provide a description of the main features associated with bottom temperature anomalies during 1999 and to put them in climatologic perspective. By anomalies we mean specifically deviations from climatology or mean seasonal cycle.

Bottom temperatures within the deeper sections of LIS exhibit extreme seasonal variations with a range of approximately 20[degrees]C. Spatial patterns of bottom temperature in LIS during summer months have distinctive features that are strongly influenced by bottom topography. Crowley (2005) has identified a cold pool within the interior of the basin as an important large temperature feature, which tends to persist throughout the spring and into the summer. The lateral distribution of temperature through the cold pool shows strong bottom fronts separating the warmer and vertically more thoroughly-mixed waters on the flanks of the channel from the colder channel waters. Bowman and Esaias (1981) have, in fact, described tidal mixing zones around the perimeter of the basin and over the several lateral sills traversing the basin. They have also described frontal regions separating the mixed from stratified waters. These mixing zones have an important influence on the response of bottom temperatures to the local surface heat flux. Anomalies in surface heat flux could be expected to have greatest influence on bottom temperature in these tidally mixed areas.

In this study we make use of hydrographic data from two regional monitoring databases: one maintained by the Connecticut Department of Environmental Protection (CTDEP) and the other by the New York Department of Environmental Protection (NYDEP). The primary database used for analyses described here consists of bottom temperatures extracted from the CTDEP database for each of the 50 stations shown in Figure 1. This data set consists of more than 4000 CTD observations. These temperature data can be extracted directly as near bottom or bottom values, and they represent a time series for bottom values that extend from 1991 to present for primary reference stations in the channel and from 1994 to present for secondary stations located generally in the channel flanks. Bottom temperature observations from NYDEP Station E10 located in the western end of LIS at 40.843[degrees]N and 73.767[degrees]W afford a significantly longer time series. The data analyzed from this station extend from 1948 through the present, although longer series for this station are available.

[FIGURE 1 OMITTED]

RESULTS AND DISCUSSION

Data from the CTDEP Stations in Figure 1 permit us to construct the surface and bottom temperature distributions using optimal interpolation for early August 1999 (Fig. 2). Within the interior of the basin to the west of the sill located between stations I2 and J2, there is evidence for the residual cold pool described by Crowley (2005) within the deeper parts of the basin. Bottom temperatures are higher within shallow, more thoroughly-mixed areas. Surface temperatures tend to be higher in the stratified areas of the basin.

[FIGURE 2 OMITTED]

Mean Annual Cycle for Bottom Temperature

A climatology for bottom temperatures at CTDEP channel stations is provided in Table 1. Monthly mean values for all stations within the interior of the basin to the west of the sill range from approximately 2[degrees]C in February to 21.5[degrees]C in September. Bottom temperatures at stations to the west of the sill have a seasonal range that is approximately 3.5[degrees]C greater than that at M3 to the east of the sill. Surface temperatures to the west of the sill have a seasonal range that is approximately 5[degrees]C greater than that at M3. This emphasizes the importance of local surface heating to the heat budget of the basin.

Bottom Temperature Anomalies

Bottom temperature anomalies for the CTDEP channel stations for the period 1991 to 2002 were estimated as the difference between monthly averaged temperatures and the climatology values in Table 1. Figure 3 shows anomaly values as a function of month and year; these values are averaged over Stations B3 through J2 within the interior of the basin. Figure 3 emphasizes that maximum monthly averaged anomalies were approximately 2.5[degrees]C. It emphasizes also that a warming trend began in 1997. Interannual variations in winter bottom temperatures can be significantly greater than those in summer. Years with elevated winter temperatures often have elevated summer temperatures as well. The year 1999 was somewhat unique because, in contrast to other years during this warming period, June, July and August had significant positive anomalies.

[FIGURE 3 OMITTED]

The total variance in the anomalies for Stations B3 through J2 decreases monotonically from west to east from approximately 1.46 [([degrees]C).sup.2] to 0.96 [([degrees]C.sup.2] Principal Component Analysis (PCA) applied to the five temperature series can identify the dominant spatial patterns associated with these anomalies. Mode 1 accounts for 83% of the total variance; its spatial structure is defined by the eigenvector in Figure 4, which shows amplification towards the western part of the basin. This structure would point to the importance of local surface heat flux in producing these anomalies rather than exchange with the coastal ocean. The principal component time series for this mode (Fig. 5) shows that 1999 was unique in having positive anomalies throughout the year.

[FIGURES 4-5 OMITTED]

To extend the analysis of bottom temperature anomalies at the CTDEP channel stations, we have estimated averages over all 50 CTDEP stations located in Figure 1 and so included data from the shallow channel flanks. Figure 6 shows time series for July and August bottom temperatures representing basin-wide averages; it emphasizes that for July and August temperatures during 1999 were well above temperatures in preceding years. Basin-wide averages during September were not, however, anomalously high.

[FIGURE 6 OMITTED]

Figure 7 shows the basin-wide averages computed for July and August partitioned by water column depth. During years with significant positive summer anomalies, there is a marked dependence on water column depth. This is especially true during the summer of 1999 when bottom temperatures at shallow stations (<20 m depth) exceeded 23.5[degrees]C. The temperatures at these stations could exceed those at deeper stations by as much as 2[degrees]C, much of this change presumably occurring across localized bottom fronts.

[FIGURE 7 OMITTED]

Bottom Temperature at NYDEP Station E10

It is useful to put the temperature variability seen at the CTDEP stations over the past decade in the context of significantly longer records available for NYDEP Station E10. Because bottom temperatures at the CTDEP channel co-vary, we have compared only CTDEP Station D3 with E10. Figures 8 and 9 show the time series for monthly-averaged bottom temperatures at E10 for the period 1948 to 2000 for August and September, respectively. For those periods of overlap, it is seen that bottom temperature fluctuations at E10 and D3 do co-vary with reasonably good agreement in magnitudes. Figures 8 and 9 emphasize also that there have been large positive anomalies in the past, exceeding those experienced since 1991.

[FIGURES 8-9 OMITTED]

This dependence of summer bottom temperature anomalies on water column depth and the amplification of anomalies from east to west indicates that local heating is a significant contributor to these anomalies. Figure 10 shows the seasonal cycle in monthly averaged local net surface heat flux anomalies from 1948 through 1999. These anomalies have been estimated from long-term insolation and meteorological observations from Brookhaven National Laboratory in Upton, New York and LaGuardia Airport. This figure suggests that there have in fact been modest increases in summer net surface heat flux over the past decade with magnitudes between 20 and 40 W/[m.sup.2]. Anomalies in the sensible heat flux dependent on air temperature represent a major contribution to these net heat flux anomalies during summer months. The bottom temperature anomalies at NYDEP Station E10, especially during August, co-vary with the heat flux anomalies in Figure 10, further strengthening the hypothesis concerning the importance of local heating.

[FIGURE 10 OMITTED]

Wind Stirring During August 1999

During 1999, bottom temperatures showed evidence of significant water column mixing between August 4 and September 1. At the shallow water stations, bottom temperatures increased only slightly, whereas those at the deeper stations increased significantly. Figure 11 shows the change in water column temperature structure at CTDEP Station D3 between August 4 and September 1 ; between August 4 and August 18 bottom temperatures increased by approximately 1.25[degrees]C, and between August 18 and September 1 bottom temperatures increased further by approximately 0.75[degrees]C. These successive increases were associated with de-stratification of the water column, which became completely mixed by September 1. Figure 12 shows a weather map on August 29, 1999; a fast moving cold front passed over LIS on this date. Wind stirring and surface cooling associated with this frontal passage made a significant contribution to vertical mixing of the water column, producing a rapid increase in bottom temperature within the deeper areas of the basin and rapid and significant increase in bottom dissolved oxygen from approximately 3 mg/L to >6.5 mg/L. This mixing event produced bottom temperatures within the deeper areas that were significantly above climatology.

[FIGURES 11-12 OMITTED]

Summer bottom and near bottom temperatures throughout LIS show a strong dependence on water column depth, with maximum temperatures occurring over the shallow flanks reflecting the extent to which the water column is vertically mixed. We conclude that year 1999 is distinctive in terms of summer bottom temperatures because of the prolonged elevated temperatures over large areas of the basin as reflected in the basin-averaged bottom temperatures during July and August. We conclude also that anomalies in net surface heat flux, principally in the sensible heat flux, made a significant contribution to these summer anomalies. Previous years did exhibit high temperatures exceeding 23.5[degrees]C in shallow waters but not for the same prolonged periods. During late August 1999, bottom temperatures in deeper waters also experienced a rapid temperature rise in response to a unique meteorologic mixing event.

ACKNOWLEDGMENTS

This research was supported by New York Sea Grant through grant # R/CE-17 to Robert Wilson and Robert L. Swanson. This publication was supported by the National Sea Grant College Program of the United States Department of Commerce's National Oceanic and Atmospheric Administration under award # NA16RG1354 to the Research Foundation of State University of New York for New York Sea Grant. The views expressed herein do not necessarily reflect the views of any of those organizations. The authors are thankful for the support of New York Sea Grant.

LITERATURE CITED

Bowman, M. J. & W. E. Esaias. 1981. Fronts, stratification and mixing and in Long Island Sound and Block Island Sound. J. Geophys. Res. 85: 2728-2742.

Crowley, H. A. 2005. The seasonal evolution of thermohaline circulation in Long Island Sound. Doctoral Dissertation, Marine Sciences Research Center, Stony Brook University, Stony Brook, NY. 217 pp.

ROBERT E. WILSON * AND ROBERT L. SWANSON

Marine Sciences Research Center, Stony Brook University, Endeavour Hall, Stony Brook, New York 11794

* Corresponding author. E-mail: [e-mail]

TABLE 1.
Monthly averaged bottom temperatures, [degrees]C

Month      B3       F3       D3       H6       J2       M3

1         3.88     4.37     4.56     4.3      4.22     5.5
2         1.84     2.07     2.02     2.19     2.42     3.56
3         2.80     2.54     2.35     2.39     2.37     3.48
4         5.33     4.44     4.15     4.29     4.57     5.1
5         9.47     7.86     7.1      7.22     8.31     8.03
6        13.57    12.32    11.96    11.91    13.31    11.82
7        16.91    16.67    16.79    16.82    18.25    16.17
8        20.41    19.98    19.86    20       20.88    18.46
9        21.56    21.43    21.37    21.39    21.77    19.64
10       18.72    19.16    19.25    19.13    18.96    17.72
11       13.36    13.99    14.6     14.06    13.39    13.16
12        8.6      9.21      9.49    9.52     9.28     9.76