1. The climate and climate variability of the wind power resource in the Great Lakes region of the United States Sharon Zhong 1 *, Xiuping Li 1, Xindi Bian 2, and Warren Heilman 2 1 Department of Geography, Michigan State University and 2 USDA Forest Service, Northern Research Station East Lansing, MI 48823 *zhongs@msu.edu Conclusions  The daily mean wind speeds exhibit a large seasonal variability, with the highest mean wind speed （ ~6.58 m s -1 ） in November through January and the lowest (~4.72 m s -1 ) in July and August.  The spatial variability of the annual mean winds is small across the entire region, and is dominated by land-water contrasts with stronger winds over the lake surface than over land.  Larger interannual variability is found during the winter months, while smaller variations occur in mid to late summer.  The interannual variability appears to have some connections to ENSO, with lower mean wind speeds and more frequent occurrences of lulls during major El Niño episodes.  Above normal ice cover of the Great Lakes appears to be associated with slightly lower wind speeds, and vice versa.  According to NARR data and the criteria established by wind energy industry, the areas over Lake Superior, Michigan, and Ontario appear to be rich in wind resources, but most land areas in the region appear to be either unsuitable or marginal for potential wind energy development. Introduction The study presents a climatology of wind at the modern wind turbine level (80-m above ground level or AGL) for the Great Lake region of the United States. The study examines the spatial distribution of mean wind across the region and the seasonal and interannual variability. The study tests the hypothesis that, similar to temperature and moisture, the characteristics of the low-level winds are modified by the thermodynamic and dynamic effects of the large water bodies in the region. The study also investigates the possible connection between the interannual variability of low-level winds across the Great Lakes region and the changes in large-scale circulation patterns associated with El Niño-Southern Oscillation (ENSO). Acknowledgements This research is supported by USDA Michigan Agricultural Experiment Station and by the USDA Forest Service Northern Research Station under agreement 07-JV-11242300-138. Results and Discussions Spatial distribution of the correlation coefficients between monthly mean wind speed and monthly MEI index (ENSO index) from 1979 to 2008. Wind speed appears to be negatively correlated with MEI. The linear trends (m s -1 a -1 ) (left column ) and the correlation coefficients (right column) for January, August and annual mean wind speeds for the 30- year period of 1979 to 2008. The numbers on top of each plot are mean, max, min and spatial standard deviation The climatology of wind power class for each grid point in the domain as determined from the 30-year averaged annual mean 80-m level wind speed from NARR. Domain-averaged monthly mean wind speed, showing large seasonal variation with weaker winds in summer and stronger winds in winter. Date Set The dataset used is the North American Regional Reanalysis (NARR ), a gridded meteorological/hydrological data set produced by National Center for Environmental Predictions (NCEP) by assimilating a large set of available observations with the NCEP mesoscale operational forecast model. NARR data set starts from Jan. 1979 and continues to present, it is dynamically consistent, and has high horizontal (32 km) and vertical (45 levels) resolution and temporal resolution (3 hourly). Of particular interest to this study is the assimilation of observed surface wind into NARR that has resulted in a significant improvement of NARR 10-m winds, and the assimilation of high resolution Great Lakes ice and temperature data that has improved NARR’s representation of climate over the Great Lakes region. For this study, we used 30-year data from 1 Jan. 1979 to 31 Dec. 2008. The domain for this study is shown below. The daily mean wind speeds over the domain are calculated by first computing wind speed from the archived zonal and meridional wind components starting at 0000 local standard time each day. Wind speeds at 80 m above ground level (AGL) were computed by simple linear interpolation in the vertical. A41F-0177 The North America domain of NARR and the topography The study domain covering the Great Lakes region. The dots are NARR grid points Spatial distribution of annual mean wind speed (m s -1 ) averaged from 1979 to 2008 over the Great Lakes region. Differences in mean wind speeds (m s -1 ) over the lakes between higher than normal and normal ice cover years (left), and between lower than normal and normal ice cover years (right). Mean winds over lakes are slightly weaker in years with higher than normal ice cover. (b) Wind speeds corresponding to different wind power classes (following Archer and Jacobson 2003). Class Speed at 80 m (m s -1 ) Suitability for wind power 1 v < 5.9 unsuitable 2 5.9 <= v < 6.9 marginal 3 6.9 <= v < 7.5 suitable 4 7.5 <= v < 8.1 suitable 5 8.1 <= v < 8.6 suitable 6 8.6 <= v < 9.4 suitable 7 v >= 9.4 suitable (a) Interannual variation and linear trend of annual mean wind speed averaged over all land points and all lake points For more detailed information abut this work, please contact Dr. Sharon Zhong at zhongs@msu.edu