Subsurface soil moisture is a primary determinant for root development and nutrient transportation in the soil and affects the tractability of agricultural vehicles. A statistical forecasting model, Vector AutoRegression (VAR), and a Long Short-Term Memory network (LSTM) were developed to forecast the subsurface soil moisture at a 20 cm depth using 9 years of historical weather data and subsurface soil moisture data from Fort Wayne, Indiana, USA. A time series analysis showed that the weather data and soil moisture have a stationary seasonal tendency and demonstrated that soil moisture can be forecasted from weather data. The VAR model estimates volumetric soil moisture of one-day ahead with an R², MAE (m³m⁻³), MSE (m⁶m⁻⁶), and RMSE (m³m⁻³) of 0.698, 0.0561, 0.0046, and 0.0382 for 2021 corn cropping season, whereas the LSTM model using inputs of previous seven days yielded R², MAE (m³m⁻³), MSE (m⁶m⁻⁶), and RMSE (m³m⁻³) of 0.998, 0.00237, 0.00002, and 0.00382, respectively as tested for cropping season of 2020 and 0.973, 0.00368, 0.00003 and 0.00577 as tested for the cropping season of 2021. The LSTM model presents a viable data-driven alternative to traditional statistical models for forecasting subsurface soil moisture.

The landscapes across much of the Midwestern United States are characterized by glacial activity that left water‐holding kettles, depressions, and potholes. Until recently, traditional watershed algorithms assumed these depressions to be errors in the elevation data and filled them as a means of correction, when many of these features may rarely fill and have the potential to dramatically affect surface flow patterns. These depressions play an important role in hydrology, water management, site planning, and agronomy. An optimized sequential depression‐filling algorithm (SDFA) was developed which fills these water‐holding features sequentially based on their respective retention capacity and contributing area. Outputs reflect the state of connectivity following the application of a user‐specified amount of rainfall excess (i.e., in excess of infiltration); depressions that have not been filled will remain as their own hydrologically common subcatchments. A performant algorithm is integral to future delivery and utilization by practitioners both in the office and in the field. The optimal set of subroutines were able to fill all of the depressions in a 239 km²watershed in northeastern Indiana in 42 s and 1.5 h on a consumer desktop computer for digital elevation models (DEMs) at 30 m and 3 m resolutions, respectively (O(n²) overall performance).