Soil Carbon Transitions (PI)
Machine‑learning estimation of land‑use transitions needed across CONUS to buffer climate‑induced soil carbon emissions.
DOE • 2023–2024Senior Member of Technical Staff · Environmental Engineer @ Sandia National Laboratories
Soil Carbon • Hydrology • Climate Extremes • Bioenergy • ESMs • Machine Learning
I develop integrated modeling approaches that fuse process-based models, Earth System Models, and machine learning to quantify climate impacts on soils, water resources, greenhouse gases, and bioenergy supply chains. My work spans field to global scales, with an emphasis on benchmarking model predictions against observations and informing decision-support tools for sustainable land management.
Based at Sandia National Laboratories (Livermore, CA), I collaborate across national labs and universities to improve representation of soil processes in large-scale models and to evaluate climate extremes, drought, and land-use transitions affecting soil organic carbon.
Carbon & nitrogen cycling, GHG fluxes, microbial processes, and observational benchmarks for model improvement.
Downscaling CMIP6, drought metrics, flood/runoff simulation, and ecohydrologic responses under climate change.
Reduced-order models, ensemble ML for SOC mapping, and LCA-linked analytics for agricultural & bioenergy cropping systems.
Machine‑learning estimation of land‑use transitions needed across CONUS to buffer climate‑induced soil carbon emissions.
DOE • 2023–2024Representing soil microbial processes in Earth System Models to evaluate impacts of climate extremes on dryland SOC dynamics.
DOE • 2023–presentBenchmarking ML projections against ESMs suggests CONUS may lose substantial SOC by 2100, guiding mitigation pathways.
ResearchWe integrate DAYCENT, DNDC, and EcoSYS to predict soil organic carbon (SOC) and crop yield across CONUS. The multi-model ensemble quantifies uncertainty and benchmarks projections for mitigation and adaptation planning.
ResearchWe fuse machine learning with process-based models to map current soil moisture and project future changes across CONUS under CMIP6 scenarios—capturing spatial patterns, trends, and uncertainty for climate-smart planning.
ResearchFast surrogates for DAYCENT/ESM outputs to predict SOC dynamics and bioenergy crop yields at policy‑relevant scales.
DevelopmentEnsemble ML improves mapping of surface SOC stocks in data‑limited northern regions for model benchmarking.
PublishedDownscaled CMIP6 indicates increased drought and extreme events across the U.S. under high‑emission scenarios.
PublishedEcohydrologic modeling for Goodwater Creek: ensemble approaches to future runoff and drought occurrence.
PublishedAssessing how soil carbon payments and multifunctional landscapes enable carbon‑negative, cost‑effective biofuels.
PublishedMost Impactful Publication Award (2024) for advancing bioenergy and environmental sustainability research.
AwardMachine‑learning & process‑based comparison for spatiotemporal ALT variation. Scientific Reports (rev., 2025).
ML surrogates for bioenergy crops. Carbon Capture Science & Technology (2025).
High‑emissions scenario analysis of future extremes. Scientific Reports (2023).
Impacts of carbon farming on sustainable aviation fuel. PNAS (2023).
Full list available on Google Scholar and CV.
For proposals or collaboration inquiries, please reach me at
sgautam@sandia.gov
Senior Member of Technical Staff, Environmental Engineer — Sandia National Laboratories