Models

Examples of how our machine learning models are able to estimate gross primary productivity and evapotranspiration.

Our models are trained using eddy covariance tower data and GOES-R advanced baseline imager (ABI) data.


June 9th 2022 ALIVE gross primary productivity gross primary productivity (µmol CO2 m-2 s-1) at 5-minute increments, showing results before and after switching out the hourly NOAA Level 2 Downward Shortwave Radiation (DSR) product for our own 5-minute DSR imagery generated from a machine learning model.


ALIVE downward shortwave radiation (DSR) at 5-minute increments, modeled from GOES-16 ABI observations using a Gradient Boosting Regression machine learning model.

July 22nd 2022


Gradient boosting model of gross primary productivity (µmol CO2 m-2 s-1) using only GOES-16 ABI variables.

June 6th 2023


Gradient boosting model of evapotranspiration derived from eddy covariance measurements. (mm per half-hour)

July 20th 2023