When Statistical Learning Violates Physics: An Extensive Mismatch in Soil Temperature-Depth Relationships in Global Maps of Soil Temperature.

Global gridded soil temperature datasets are important to understand and explain spatial patterns and processes in many life and environmental sciences, but products based on in situ measurements are still available to a very limited extent. Global m…

Preview only. Read the full abstract at the source