Station / Map Differences

[qjhart]

We have differences between the Spatial CIMIS prediction as the Stations, and the Station data itself. We have two plans for dealing with that.

Quinn's Plan

One thing we can do is to take the long term station data, and calculate the FFT transform using that data. Then we have some additional information, we have the P0 and P1 powers from the stations. We can then calculate two ratios P0_station / P0_map and P1_station / P1_map, and then calculate a spline fit of these two rations for the state. This allows us a better method to see where the differences are and what parts of the state they effect.

We can then apply those ratios to the map data and compare the final data from that to the map station data alone.

Bekele's Plan

Bekele thinks an alternative would be to take the long term data, and use that as another set of inputs into the ETO calculation. This is more elaborate since it would require that we calculate 52 new dates, and then run the ETO calculation on all of them. In addition, we'd need to come up with some method of taking the station data Rs and using that instead of the Spatial CIMIS Rs. I'm not sure of the best method for that. We'd have to somehow match the K factor at the stations to the K factor from the GOES data, and that we can't spline fit through.

We could only use the best possible fit between them. But that wouldn't help us find spatial differences between the two for that component.