StruMPL: Multi-task Dense Regression for Forest AGB Estimation

There's this interesting new method called StruMPL that's designed to estimate forest aboveground biomass (AGB) using data from Earth observations. It faces a challenge because it relies on two different types of data that don’t quite match up. On one hand, spaceborne lidar gives us canopy structure info from millions of locations, but it doesn't provide biomass estimates. On the other hand, ground-based plots do offer biomass data, but they’re often biased and lack structural details. StruMPL creatively addresses this by treating it as a multi-task regression problem, managing the tricky missing data and applying known allometric laws. You can find more details about the research on arXiv.