National Center for Ecological Analysis and Synthesis 02-15-2008 raster digital data Science \\ebm\homes\data\marine_threats\data\completed\impacts\transformed\sst\grid\sst Science 02-15-2008 sst Changes in sea surface temperature as a result of climate change will have varying effects on species and ecosystems depending on background temperature variation at a location. Potential for ecological change in response to changes in sea surface temperature can be measured as the frequency of temperature anomalies, where the temperature exceeds a threshold value like the long-term mean, or by a measure of the magnitude (°C) of the anomalies themselves (S19, 20). We used 4.6 km (nominally 21 km2 at the equator) Advanced Very High Resolution Radiometer (AVHRR) Pathfinder Version 5.0 SST data produced by NOAA's National Oceanographic Data Center and the University of Miami's Rosenstiel School of Marine and Atmospheric Science (http://pathfinder.nodc.noaa.gov) to create a global database of temperature anomalies. This database was then used to calculate differences in anomaly frequency between 2000-2005 and 1985-1990. We first developed a climatology or long-term weekly average for each approximately ~4 km grid cell using data from 1985 to 2005. These data provide a baseline for determining when temperatures are unusually warm. We then developed two metrics for quantifying climate change as a driver of change in marine ecosystems. For our first metric, we calculated the number of times the SST anomaly exceeded the standard deviation of SSTs for that location and week of the year. This threshold-based approach accounts for natural variability at a given location, which can vary widely from place to place, by incorporating the standard deviation. The second metric was the mean SST anomaly temperature value normalized by the standard deviation, a continuous variable approach that also accounts for differences in variability from region to region. The resolution of the temperature data is coarser (21 km2) than the resolution of analysis (1 km2); each 1 km2 cell within the larger cells was assigned the value of the two metrics for the 21 km2 cell. 11 These two metrics of SST should be correlated, and indeed they were highly correlated for a test region (45°N-22.5°N, 90°W-67.5°W) in the Western Atlantic and Caribbean that has a wide range of values (linear regression; R2 = 0.91, P < 0.00001) and for a random sample of 20,000 cells around the globe (linear regression; R2 = 0.998, P < 0.00001; these sub-samples were necessary because the data layers are too large to analyze in their entirety). Consequently, we focused on only the threshold anomaly metric as a measure of SST change for our analyses. We then developed a change metric by subtracting the number of non-zero positive anomalies in the early period (1985-1990) from the number in the recent period (2000-2005); these are the data included in our cumulative impact model. To test for sensitivity of results to our choice of focal years (2000-2005), we used the Western Atlantic and Northern Caribbean as a test region to compare values for the past 10 years (1995-2005) and the entire dataset (1985-2005) for both metrics describe above. If mean and variance values are increasing over time, then one should see decreasing fits in the data with an increasing span of time being compared. Indeed, all comparisons were highly significant (P<0.00001) with high but decreasing R2 values when the most recent data were compared to the entire dataset, the past 10 years, and the most recent period (linear regression; anomalies 5 yrs vs. 10 yrs R2 = 0.961, 5 yrs vs. 20 yrs R2 = 0.93; normalized means 5 yrs vs. 10 yrs R2 = 0.9997, 5 yrs vs. 20 yrs R2 = 0.9993). Our approach did not account for potential presence of thermo-tolerant species or local-scale oceanography that could modify how resistant a local ecosystem is to changes in SST. What happens in the vast stretches of the world's oceans - both wondrous and worrisome - has too often been out of sight, out of mind. The sea represents the last major scientific frontier on planet earth - a place where expeditions continue to discover not only new species, but even new phyla. The role of these species in the ecosystem, where they sit in the tree of life, and how they respond to environmental changes really do constitute mysteries of the deep. Despite technological advances that now allow people to access, exploit or affect nearly all parts of the ocean, we still understand very little of the ocean's biodiversity and how it is changing under our influence. The goal of the research presented here is to estimate and visualize, for the first time, the global impact humans are having on the ocean's ecosystems. Our analysis, published in Science February 15, 2008, shows that over 40% of the world's oceans are heavily affected by human activities and few if any areas remain untouched. Please see supplimental online materials for more information. en 02-15-2008 publication date Complete As needed -180.000000 180.000000 90.000000 -90.000000 -18040095.19613218040134.245271-9020047.3480709020067.372632 REQUIRED: Reference to a formally registered thesaurus or a similar authoritative source of theme keywords. impact marine global modeling global ocean marine current 2007 2008 Data available from NCEAS is distributable (please cite correctly). Please contact before distributing dirived data. For original data, please visit http://www.nceas.ucsb.edu/GlobalMarine For distribution of orignal data, please use proper citation. Ben Halpern National Center for Ecological Analysys and Synthesis Assistant Research Scientists mailing and physical address

735 State Street

Santa Barbara California 93101 USA (805)-892-2531 (805)-892-2510 halpern@nceas.ucsb.edu National Center for Ecologial Analysis and Synthesis Microsoft Windows 2000 Version 5.2 (Build 3790) Service Pack 2; ESRI ArcCatalog 9.2.0.1324 Raster Dataset Raster Grid Cell 19305 38610 1 934.478877934.47887732Upper LeftFALSEDefault1matrix valuesTRUE row and column 934.478877 934.478877 meters Mollweide0.0000000.0000000.000000 D_WGS_1984WGS_19846378137.000000298.257224 GCS_WGS_1984WGS_1984_Mollweide National Center for Ecoogical Analysys and Synthesis http://www.nceas.ucsb.edu/GlobalMarine Downloadable Data -1241.858 -1241.858 20080213 National Center for Ecological Analysis and Synthesis Colin Ebert GIS Analyst REQUIRED: The mailing and/or physical address for the organization or individual. REQUIRED: The city of the address. REQUIRED: The state or province of the address. REQUIRED: The ZIP or other postal code of the address. 805-892-2500 ebert@nceas.ucsb.edu http://www.nceas.ucsb.edu/GlobalMarine/ FGDC Content Standards for Digital Geospatial Metadata FGDC-STD-001-1998 local time http://www.esri.com/metadata/esriprof80.html ESRI Metadata Profile enhttp://www.esri.com/metadata/esriprof80.htmlESRI Metadata Profile {8E1657B8-B890-4B85-A3C0-EAB6FB3CB6C0}2008021310112800FALSE20080213152809002008021315280900Metadata imported.C:\DOCUME~1\Ebert\LOCALS~1\Temp\xml6D.tmp2008021310112800Microsoft Windows 2000 Version 5.2 (Build 3790) Service Pack 2; ESRI ArcCatalog 9.2.0.1324sst-18040095.19613218040134.2452719020067.372632-9020047.348071-18018090-901ISO 19115 Geographic Information - MetadataDIS_ESRI1.0datasetfile://\\ebm\homes\data\marine_threats\data\completed\impacts\transformed\sst\grid\sstLocal Area Network002-1241.858Raster DatasetWGS_1984_Mollweide2138610934.4789Meter1 Meter = 1 Meter(s)19305934.4789Meter1 Meter = 1 Meter(s)20080213