Sagar Gautam

. Representing soil organic carbon (SOC) dynamics in Earth system models (ESMs) is a key source of uncertainty in predicting carbon climate feedbacks. Machine learning models can help identify dominant environmental controllers and their functional relationships with SOC stocks. The resulting knowledge can be implemented in ESMs to reduce uncertainty and better predict SOC dynamics over space and time. In this study, we used a large number of SOC field observations (n = 54,000), geospatial datasets of environmental factors (n = 46), and two machine learning approaches (Random Forest (RF) and Generalized Additive Modeling (GAM)) to: (1) identify dominant environmental controllers of global and biome-specific SOC stocks, (2) derive functional relationships between environmental controllers and SOC stocks, and (3) compare the identified environmental controllers and predictive relationships with those in Coupled Model Intercomparison Project phase six (CMIP6) models. Our results showed that diurnal temperature, drought index, cation exchange capacity, and precipitation were important observed environmental controllers of SOC stocks. RF model predictions of global-scale SOC stocks were relatively accurate (R2 = 0.61, RMSE = 0.46 kg m−2). In contrast, precipitation, temperature, and net primary productivity explained > 96 % of ESM-modeled SOC stock variability. We also found very different functional relationships between environmental factors and SOC stocks in observations and ESMs. SOC predictions in ESMs may be improved significantly by including additional environmental controls (e.g., cation exchange capacity) and representing the functional relationships of environmental controllers consistent with observations.

Long-term hydrologic data are required to quantify the impacts of management and climate change on runoff at the field scale where management practices are applied. This study was conducted to evaluate the impacts of long-term management and climate change on runoff from a small watershed managed with no-till (NT) and grazed pasture systems. Further, study was conducted to evaluate the performance of the two widely used hydrologic models; Agricultural Policy Environmental eXtender (APEX) and Soil and Water Assessment Tool (SWAT) for their ability to simulate stream flow. In this study, two different calibration approaches, the parameter estimation (PEST) and sequential uncertainty domain parameter fitting (SUFI-2) method were employed for model evaluation and calibration of APEX and SWAT models, respectively. The specific objectives of the study were to: i) to calibrate the APEX model using an inverse modeling approach (PEST-APEX) (ii) simulate the impacts of cropping management, grazing management, climate change, and tillage system on runoff, and (iii) compare simulation results generated by two widely used hydrologic models, APEX and SWAT. The study was conducted using four small watersheds (one no-till and three grazed pasture watersheds) located in the North Appalachian Experimental Watersheds (NAEW) near Coshocton, Ohio and one larger watershed; Millcreek watershed located in Coshocton and Holmes counties of Ohio. Data show that the use of PEST-APEX and SUFI-2 resulted in efficient calibration of the models compare to trial and error methods.The results also indicate that the corn-soybean-rye rotation under no-till system is significantly beneficial for controlling runoff compared to continuous corn. The climate change scenarios indicate that runoff from the no-till watershed is the most sensitive to the precipitation, and interactions of precipitation, temperature, and carbon dioxide concentrations. Results from grazed pasture watershed study demonstrate the benefits of hayed meadow over grazed pasture and conclude that surface runoff is affected by soil properties, and can be reduced by using buffer strips of perennial grasses at the downslope of the watershed. Further, the results from the Millcreek watershed study indicate that the SWAT can simulate the stream flow reasonably well compared to the APEX due to better ground water parameterization and finer subdivision of the watershed

Understanding the influence of environmental factors on soil organic carbon (SOC) is critical for quantifying and reducing the uncertainty in carbon climate feedback projections under changing environmental conditions. We explored the effect of climatic variables, land cover types, topographic attributes, soil types and bedrock geology on SOC stocks of top 1 m depth across conterminous United States (US) ecoregions. Using 4559 soil profile observations and high-resolution data of environmental factors, we identified dominant environmental controllers of SOC stocks in 21 US ecoregions using geographically weighted regression. We used projected climatic data of SSP126 and SSP585 scenarios from GFDL-ESM 4 Earth System Model of Coupled Model Intercomparison Project phase 6 to predict SOC stock changes across continental US between 2030 and 2100. Both baseline and predicted changes in SOC stocks were compared with SOC stocks represented in GFDL-ESM4 projections. Among 56 environmental pr...

National scale projections of bioenergy crop yields and their environmental impacts are essential to identify appropriate locations to place bioenergy crops and ensure sustainable land use strategies. In this study, we used the process‐based Daily Century (DAYCENT) model with site‐specific environmental data to simulate sorghum (Sorghum bicolor L. Moench) biomass yield, soil organic carbon (SOC) change, and nitrous oxide emissions across cultivated lands in the continental United States. The simulated rainfed dry biomass productivity ranged from 0.8 to 19.2 Mg ha−1 year−1, with a spatiotemporal average of  Mg ha−1 year−1, and a coefficient of variation of 35%. The average SOC sequestration and direct nitrous oxide emission rates were simulated as  Mg CO2e ha−1 year−1 and  Mg CO2e ha−1 year−1, respectively. Compared to field‐observed biomass yield data at multiple locations, model predictions of biomass productivity showed a root mean square error (RMSE) of 5.6 Mg ha−1 year−1. In com...

Various approaches of differing mathematical complexities are being applied for spatial prediction of soil properties. Regression kriging is a widely used hybrid approach of spatial variation that combines correlation between soil properties and environmental factors with spatial autocorrelation between soil observations. In this study, we compared four machine learning approaches (gradient boosting machine, multinarrative adaptive regression spline, random forest, and support vector machine) with regression kriging to predict the spatial variation of surface (0–30 cm) soil organic carbon (SOC) stocks at 250-m spatial resolution across the northern circumpolar permafrost region. We combined 2,374 soil profile observations (calibration datasets) with georeferenced datasets of environmental factors (climate, topography, land cover, bedrock geology, and soil types) to predict the spatial variation of surface SOC stocks. We evaluated the prediction accuracy at randomly selected sites (v...

Accurate representation of environmental controllers of soil organic carbon (SOC) stocks in Earth System Model (ESM) land models could reduce uncertainties in future carbon–climate feedback projections. Using empirical relationships between environmental factors and SOC stocks to evaluate land models can help modelers understand prediction biases beyond what can be achieved with the observed SOC stocks alone. In this study, we used 31 observed environmental factors, field SOC observations (n = 6,213) from the continental United States, and two machine learning approaches (random forest [RF] and generalized additive modeling [GAM]) to (a) select important environmental predictors of SOC stocks, (b) derive empirical relationships between environmental factors and SOC stocks, and (c) use the derived relationships to predict SOC stocks and compare the prediction accuracy of simpler model developed with the machine learning predictions. Out of the 31 environmental factors we investigated...

ABSTRACT The Agricultural Policy Environmental eXtender (APEX), a comprehensive hydrologic model well-suited for small watersheds, requires understanding of the input parameters for improved calibration. The “trial and error” method for calibrating the APEX model has been used very commonly in previous studies. In this study, the automatic calibration software Parameter Estimation (PEST) was combined with the conventional trial-and-error method to improve APEX calibration. The proposed Combined PEST and Trial–Error (CPTE) approach can overcome: (i) weaknesses of “Trial–Error” method in terms of tediousness and subjectivity involved in the decision to end a calibration, and (ii) drawback of PEST in that it may lead to biased simulation due to ignoring local specific condition. A case study was developed to verify the CPTE approach. The results based on APEX runoff simulation indicate that the CPTE approach greatly improved the calibration of APEX model with respect to model performance criteria. Coupling inverse modeling and trial–error manual method can be an efficient and effective alternative in calibrating the APEX model.