Water returned to the atmosphere as evapotranspiration (ET) is approximately 1.6x greater than global river discharge and has wide-reaching impacts on groundwater and streamflow. In the U.S. Midwest, widespread land conversion from prairie to cropland has altered spatiotemporal patterns of ET, yet there is no consensus on the direction of change in ET or the mechanisms controlling changes. We aimed to harmonize findings about how land use change affects ET in the Midwest. We measured ET at three locations within the Long-Term Agroecosystem Research (LTAR) network along a latitudinal gradient with paired rainfed cropland and prairie sites at each location. At the northern locations, the Upper Mississippi River Basin (UMRB) and Kellogg Biological Station (KBS), the cropland has annual ET that is 84 and 29 mm/year higher, respectively, caused primarily by higher ET, likely from soil evaporation during springtime when agricultural fields are fallow. At the southern location, the Central Mississippi River Basin (CMRB), the prairie has 69 mm/year higher ET, primarily due to a longer growing season. To attribute differences in springtime ET to specific mechanisms, we examine the energy balance using the Two-Resistance Method (TRM). Results from the TRM demonstrate that higher surface conductance in croplands is the primary factor leading to higher springtime ET from croplands, relative to prairies. Results from this study provide critical insight into the impact of land use change on the hydrology of the U.S. Corn Belt by providing a mechanistic understanding of how land use change affects the water budget.

Nitrous oxide (N2O) is one of the important greenhouse gases (GHGs), with its global warming potential 265 times greater than that of carbon dioxide (CO2). About 60% of the anthropogenic N2O emission is from agriculture production. To date, estimating N2O emissions from cropland remains a challenging task because the related microbial origin processes (e.g. incomplete nitrification and denitrification) are controlled by a diverse factors of climate, soil, plant and human activities. In this study, we developed a ML model with physical/biogeochemical domain knowledge, namely knowledge guided machine learning (KGML), for simulating daily N2O fluxes from the agriculture ecosystem. The Gated Recurrent Unit (GRU) was used as the basis to build the model structure. A range of ideas have been implemented to optimize the model performance, including 1) hierarchical structure based on variable causal relations, 2) intermediate variable (IMV) prediction and transfer, 3) inputting IMV initials for constraints, 4) model pretrain/retrain, and 5) multitask learning. The developed KGML was pre-trained by millions of synthetic data generated by an advanced PB model, ecosys, and then re-trained by observations from six mesocosm chambers during three growing seasons. Six other pure ML models were developed using the same data from mesocosm chambers to serve as the benchmark for the KGML model. The results show that KGML can always outperform the PB model in efficiency and ML models in prediction accuracy of capturing N2O flux magnitude and dynamics. Besides, the reasonable predictions of IMVs increase the interpretability of KGML. We believe the footprint of KGML development in this study will stimulate a new body of research on interpretable machine learning for biogeochemistry and other related geoscience processes.

Carbon monoxide (CO) is an ozone precursor, oxidant sink, and widely-used pollution tracer. The importance of anthropogenic versus other CO sources in the US is uncertain. Here we interpret extensive airborne measurements with an atmospheric model to constrain US fossil and non-fossil CO sources. Measurements reveal a low bias in the simulated CO background and a 30% overestimate of US fossil CO emissions in the 2016 National Emissions Inventory. After optimization we apply the model for source partitioning. During summer, regional fossil sources account for just 9-16% of the sampled boundary layer CO, and 32-38% of the North American enhancement-complicating use of CO as a fossil fuel tracer. The remainder predominantly reflects biogenic hydrocarbon oxidation plus fires. Fossil sources account for less domain-wide spatial variability at this time than non-fossil and background contributions. The regional fossil contribution rises in other seasons, and drives ambient variability downwind of urban areas.