Hello
How would you approach generating spatially averaged surface dataset over a catchment of interest? I mean the same setup as single-point cases, the only difference, the surface data is aggregated over a catchment.
The first idea that comes to mind is to subset the global surface dataset (like for single-point cases), selecting grid cells that overlap the catchment. Then taking the mean of all surface data variables over these cells. But I believe it is incorrect to simply average variables like the fractional saturated area. My tests show significant degradation in model performance when doing so.
Another idea is to subset the surface data at a given lat,lon (e.g., centroid) in the catchment. Then manually calculate and replace the values of variables where data is available. I am not sure which variables to change and how to calculate them.
Could you please let me know what you think and if there are examples/scripts that could be helpful to do this.
Thank you
How would you approach generating spatially averaged surface dataset over a catchment of interest? I mean the same setup as single-point cases, the only difference, the surface data is aggregated over a catchment.
The first idea that comes to mind is to subset the global surface dataset (like for single-point cases), selecting grid cells that overlap the catchment. Then taking the mean of all surface data variables over these cells. But I believe it is incorrect to simply average variables like the fractional saturated area. My tests show significant degradation in model performance when doing so.
Another idea is to subset the surface data at a given lat,lon (e.g., centroid) in the catchment. Then manually calculate and replace the values of variables where data is available. I am not sure which variables to change and how to calculate them.
Could you please let me know what you think and if there are examples/scripts that could be helpful to do this.
Thank you