search

EcoExtreML / STEMMUS_SCOPE

Integrated code of SCOPE and STEMMUS
https://EcoExtreML.github.io/STEMMUS_SCOPE
GNU General Public License v3.0
14 stars 6 forks source link

Run the model at global scale #18

Open SarahAlidoost opened 2 years ago

SarahAlidoost commented 2 years ago

Currently, the model runs at the site scale. Can we make the model global?

yijianzeng commented 2 years ago

Hi, Sarah, i think this question is a general one to all? In short, yes. Although it is running at site scale, STEMMUS-SCOPE can be ran at global scale, provided that we have all relevant input and forcing data.

For example, we are now running at 170sites, imagine we have a million sites on a global scale ... nr is not exact, but just for an example

bobzsu commented 2 years ago

The idea was to run it on global scale driven by ERA5 data. We also need to identify the PFT on a land use map and soil map.

SarahAlidoost commented 2 years ago

The current code of the model works with point data. To run it globally, it would be ideal if the code works with raster data (map). For now, we can assume that the simulations at one pixel are independent of other pixels. But eventually, there will be a need for communication between pixels too. Am I right?

bobzsu commented 2 years ago

At the moment, STEMMUS-SCOPE is one dimensional, meaning no exchanges among grids even if we run it for grids. The only consideration is that if there is too much precipitation on a grid (generating surface runoff), this is not kept. The solution is to call the coupled STEMMUS-TeC that closes water budget on a catchment basis. We need to discuss how far we want to go.