Great Work for GW, GCC.

I have an interest - in 2013 wind farm capacities in many midwest locations lowered from 2011 and 2012 values (I have a paper on wind energy in my research bench area here on research grate).  So I am interested in comparing 2011-2013 large scale wind vectors in the american midwest.

Are the wind velocity vectors original observations with no adjustment?

Have you felt that the wind speeds on the northern continents are becoming weaker with time? May be you could use the data to look at if there is a global stilling over the past decades.

Something wrong with the computer program, Dudley? It should be the westerly in the southern oceans for almost all months, but I found the most frequent wind is easterly there.

Thanks for noticing that. I will check it out. The GSOD database doesn't contain wind direction and this is the first time I've done anything with the wind data. In order to get a wind direction I first calculate the pressure field and then set the wind direction to minus the gradient of the pressure field.

you can find a deviation of the thermal wind here

http://homework.uoregon.edu/pub//class/atm/transport.html

http://homework.uoregon.edu/pub//class/atm/thermal.html

After looking over the velocity vectors I have made a few changes. I now relax (smooth) the pressure field and have decreased the grid size from 5 deg to 2 deg and am regenerating all of the pressure contours. I also forgot to account for the cos(latitude) in the pressure gradient so I am also regenerating all of the velocity vectors. The rain data is so sparse that I'm going to have to do something else with it too. While almost all of the 11,000 GSOD stations consistently report temperature, fewer consistently report wind, and even fewer report rainfall. I worked at TVA for 13 years and I know they had to adjust rainfall data based on the elevation changes of the reservoirs, so the rainfall data may simply be more problematic. I personally would never measure rainfall with one of those silly little plastic funnels. A swimming pool would be a much better way to measure rainfall. If the wind is blowing, it's a wonder the little funnels catch anything at all. I've seen ours indicate a half inch of rain or less while a wheelbarrow sitting next to it got six inches of water.

large aperture rainfall collectors are more susceptible to evaporative loss - particularly on partial cloudy days with some rainfall - so there is a balance between surface area, accuracy, and evaporation.   Personally, I think the main problem with precip data is that we don't sample it at high spatial resolution

It is grandiose work!!!

Interesting to see is there change of circulation epochs in the 90th (change of zonal transport on meridional transport) in the northern hemisphere?