EC-GPR-TCFs

Workflow for upscaling terrestrial carbon fluxes (TCFs) from a GPR model trained with LAI and climate variables at eddy-covariance towers.

Author: Pablo Reyes-Muñoz

Code: Pablo Reyes-Muñoz

Workflow for upscaling TCFs from MCD15A3H and ERA-5 through Google Earth Engine, from the paper "Tower-to-global upscaling of terrestrial carbon fluxes driven by Copernicus ERA-5 and MODIS-LAI data".

1. Please, download the source code available in this site (TCFs_upscaling / main).



The code is formed by a set of Python functions



2. The compose_image function format the predictors input for the GPR function below


3. Calculate_Green function is the core of the GPR algorithm implemented in GEE


4. The map_loop function iterates over the defined temporal windows





For training and exporting customized models from ARTMO to GEE , please follow the guidelines in https://github.com/msalinero/ARTMOtoGEE.git

A workflow in Python to produce time series mapping over a region of interest can be found in this link