I wonder if there is a correlation coefficient or parameter difference between the land surface and atmospheric temperature. The reason for this is to obtain information from the atmospheric temperature to calculate the land surface temperature in rainy seasons using remote sensing.
If one cannot measure the temperature due to clouds, one indeed needs to model it. See e.g. http://www.gmes-geoland.info/
A simplified approach some people use is to link USSA (United States Standard Atmosphere) atmospheric temperature profiles to ground temperature. They re-scalate the temperature profiles with the ground temperature and you might do the opposite. But from your question it's not clear whether you have a measured temperature profile.
In remote sensing of an atmospheric profile, one needs an observation error and a background error. The latter can be complex, including error correlation between surface temperature and other variables of interest to the retrieval process such as temperature and humidity. These error correlations can be estimated locally from ensembles of forecasts. This is what is currently done in operationl data assimilation for weather prediction. Understanding that this framework is only availale in NWP centers, one could use at least level-to-level air temperature and skin temperature error correlation estimates.
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing