Wetting and Drying of Soil: From Data to Understandable Models for Prediction

摘要

Soil moisture is critical to agriculture, ecology, and certain natural disasters. Existing soil moisture models often fail to predict soil moisture accurately for time periods greater than a few hours. To tackle this problem, we introduce in this paper two novel models, the Naive Accumulative Representation (NAR) and the Additive Exponential Accumulative Representation (AEAR). The parameters in these models reflect hydrological redistribution processes of gravity and suction. We validate our models using soil moisture and rainfall time series data collected from a steep gradient post-wildfire site in Southern California. Data analysis is challenging, since rapid landscape change in steep, burned hillslopes is typically observed in response to even small to moderate rain events. We found that the AEAR model fits the data well for three distinct soil textures at different depths below the ground surface (at 5cm, 15cm, and 30cm). Similar strong results are demonstrated in controlled soil moisture experiments. Our recommended AEAR model has been validated as effective and useful by earth scientists, giving better forecasts than existing models for time horizons of 10 to 24 hours.