Gulf Stream Sea Surface Temperature Shi1 Peter Willauer ’15 Environmental Studies Program Colby College, Waterville, ME Abstract Discussion Sea surface temperatures (SST) of the Gulf Stream change enormously on year to The Advanced Very High ResoluHon Radiometer measures the earth in 5 wide year basis. In this study, however, I invesHgated the trends and temperature shiIs spectral bands and is able to gather Sea Surface temperature around the over a 20-year period determining if a clear shiI is evident in this parHcular world. AVHRR uses thermal infrared remote sensing to measure the “skin current. I found no significant trends, but I did find several interesHng results. I SST” at a depth of 10 µm (Maurer, 2002). Figure E (below) shows how sea used ArcGIS to gather and analyze data from NOAA’s satellite (AVHRR) and surface temperature represents such a small amount of the ocean developed an Ordinarily Least Squares regression model comparing a monthly temperatures. average over 20 years, to determine if the temperature shiIs over Hme were not random. I was able to isolate slopes with significant P-values (P < .05) and Figure E determine a trend of those significant values. IntroducHon Gulf Stream shiIs have a profound impact on the weather throughout the North AtlanHc and surrounding land. The warm water, carried north from the tropics, carries warm air north, severely influencing weather generated in the troposphere. Slight shiIs in this massive ocean current have more effect on the weather and climate than previously thought (Blackmon, 1993). As technology improves and satellites and radiometers become more accurate, and real Hme sea surface temperatures become easier to measure and shiIs become more apparent. As one of the most widely studied variables in oceanography, SST accuracy is an improvement for understanding climate change and weather paerns , so it was Figure X crucial to obtain reliable, organized data from NOAA’s very high resoluHon There has been significant research on documenHng the accuracy of the AVHRR radiometer (Emery, 2001). due to its sensiHvity regarding water vapor and the effect on results. Figure C(i) Figure C (ii) Technology currently exploring ways to improve this problem involve algorithms compensang for the error, but buoys and ships gathering data sHll remain the most accurate way to compare and confirm temperature trends. Methods (Vasquez, 1998) I used ArcGIS 10 to import data from NOAA’s Pathfinder server and when trends The variaon in technology over Hme can be difficult to navigate around, did not seem apparent from year to year, I used R-Script to perform an Ordinarily especially when discussing monthly averages. Water vapor from one day can Least Squares regression on 20 monthly averages. This consisted of gathering data skew a monthly average significantly, thus compromising the integrity of the from individual pixels from 1986-2006, deriving the slopes of change by graphing study. The AVHRR’s improved sensiHvity makes the daily SST readings them, then compuHng a best fit line. (see figure Ci-ii) extremely accurate and useful for real Hme endeavors and weather predicHons, but compuHng trends in such a rapidly changing environment is less accurate. I determined which pixels had a 95% probability of not being random( P < .05), More data and analysis is needed to validate the results in this study because creang a refined map of only significant slopes. From here, I isolated what percent there are many variables unaccounted for. of the significant slopes were posiHve and negave. The graphs above show the temperatures of individual pixels over a 20 year period taken from areas of high and low slopes. Figure C(i) shows posiHve slopes (increasing temperature) and Figure C(ii) shows negave slopes (decreasing temperature). Conclusion Results The results indicate a general increase in Sea Surface Temperature of the waters surrounding the Gulf Stream, especially in Northern areas. In a similar, but more IniHally, no correlaon existed for the slopes of individual pixels over the 20 year extensive study performed by NOAA which released in April 2013, indicates ocean period (Figure A), but the regression analysis revealed a trend among the significant temperatures off the northeast coast in 2012 were the warmest they’ve been in slopes. In Figure B, several pockets of significant slopes are evident throughout the 150 years, suggesHng the changes in the Gulf Stream represent more than simply North AtlanHc. I calculated which areas had posiHve and negave slopes. Figure D shiIing weather trends (NASA 2013). shows the posiHve and negave slopes of only the significant slopes. The regression analysis of the individual pockets shows posiHve and negave trends over 20 years. The slopes are not parHcularly strong, but are stasHcally References hbp://www2.hawaii.edu/~jmaurer/sst/ significant. (Figure C(i-ii)). The results indicate that 84 % of these significant slopes are posiHve, displaying an increase in temperature over 20 years (Figure D). Figure A Figure B hbp://resources.esri.com/help/9.3/ArcGISDesktop/com/Gp_ToolRef/Spaal_StasHcs_toolbox/ regression_analysis_basics.htm hbp://woce.nodc.noaa.gov/woce_v3/wocedata_2/sat_sst/avhrr/docs/usr_gde4_0.htm#2.2 hbp://science.nasa.gov/science-news/science-at-nasa/2004/05mar_arcHc/ hbp://www.nefsc.noaa.gov/press_release/2013/SciSpot/SS1304/ Emery, W. J., et al. "EsHmang sea surface temperature from infrared satellite and in situ temperature data." BULLETIN-AMERICAN METEOROLOGICAL SOCIETY 82.12 (2001): 2773-2786. BLACKMON, MAURICE L. "Surface climate variaons over the North AtlanHc Ocean during winter: 1900-1989." J. Climate 6 (1993): 1743-1753. Figure D (above) shows significant slopes of change from 1986-2006. Red indicates an average In Figure B above, the pockets of turquoise In figure A (above) posiHve slopes represent increase, while blue indicates an average decrease. Acknowledgements represent a high P-value, meaning a high pockets where temperature generally probability of the informaon having increased over Hme (shown in orange) while Special thanks goes to Manny Gimond for his inspiraonal atude and incredible knowledge of GIS , and Philip significant trends. negave slopes represent where temperature Nyhus for constant support throughout this study. generally decreased. .