Improving Weather Forecasts: A Tale of Soil Moisture Data

Scientists are diving into the world of data assimilation to boost weather and hydrological forecasts. They've combined a land data assimilation component with a global atmospheric model called NICAM-LETKF. This mix allows them to conduct two types of experiments: strongly and weakly coupled land-atmosphere data assimilation. What's the big deal? Assimilating soil moisture (SM) data helps improve the analysis and forecasts of atmospheric variables. It even reduces a warm bias in the lower troposphere where dry soil moisture exists. But here's where it gets tricky—updating soil moisture by assimilating atmospheric observations can backfire due to strange error connections between atmospheric observations and land model variables. Strongly coupled data assimilation shows its power in regions like the Sahel and equatorial Africa during the rainy season, from May to October. These areas benefit from updates in atmospheric variables through soil moisture data assimilation, especially when rainfall increases. Interestingly, these hotspots align with areas previously identified where initializing global soil moisture boosts the prediction of seasonal rainfall.