The statewide average solar radiation correlations we had for our experiment were:
Air Temperature: 0.363050199
Relative Humidity: -0.491531947
Soil Temperature: 0.160029459
When calculating correlations, values that are close to +1 or -1 indicates that the data sets are well correlated. Unfortunately our values are closer to zero, which means that solar radiation is not well correlated with air temperature, relative humidity, and soil temperature. Overall, relative humidity had the best correlation and soil temperature had the worst when compared to solar radiation.
There are a number of things that could be done to improve the overall results of a similar project in the future. One way to get more accurate statistics is to use more data. That can mean more years of data, or data from more stations.
In addition, when we pulled all observations we were not able to easily see what was good or bad data until we graphed it. In some cases, we had to completely throw out a station because of the amount of bad data. If we were to use quality controlled data in the future, then we could easily neglect the bogus observations individually without having to leave out an entire station.
Since solar radiation is not measured at very many stations across the state, it might also be worthwhile to compare estimated solar radiation to weather measurements at stations other than ECONET.