Comprehensive land models are subject to significant parametric uncertainty, which can be hard to quantify due to the large number of parameters and high model computational costs. We constructed a large parameter perturbation ensemble (PPE) for the Community Land Model vesion 5.1 with biogeochemistry configuration (CLM5.1-BGC). We performed more than 2000 simulations perturbing 211 parameters across six forcing scenarios. This provides an expansive dataset, which can be used to identify the most influential parameters on a wide range of output variables globally, by biome, or by plant functional type. We found that parameter effects can exceed scenario effects and that a small number of parameters explains a large fraction of variance across our ensemble. The most important parameters can differ regionally and also based on the forcing scenario. The software infrastructure developed for this experiment has greatly reduced the human and computer time needed for CLM PPEs, which can facilitate routine investigation of parameter sensitivity and uncertainty, as well as automated calibration.

Plant stomata mediate the fluxes of both carbon and water between the land and the atmosphere. The ratio between photosynthesis and stomatal conductance (gs), or intrinsic water-use efficiency (iWUE), can be directly inferred from leaf or tree-ring carbon isotope composition. In many Earth system models, iWUE is inversely proportional and controlled by a parameter (g1M) in the calculation of gs. Here we examine how iWUE perturbations, setting g1M to the 5th (low) and 95th (high) percentile for each plant type based on observations, influence photosynthesis using coupled Earth System model simulations. We find that while lower iWUE leads to reductions in photosynthesis nearly everywhere, higher iWUE had a photosynthetic response that is surprisingly regionally dependent. Higher iWUE increases photosynthesis in the Amazon and central North America, but decreases photosynthesis in boreal Canada under fixed atmospheric conditions. However, the photosynthetic response to higher iWUE in these regions unexpectedly reverses when the atmosphere dynamically responds due to spatially differing sensitivity to increases in temperature and vapor pressure deficit. iWUE also influences the photosynthetic response to atmospheric CO2, with higher and lower iWUE modifying the total global response to elevated 2x preindustrial CO2 by 6.4% and -9.6%, respectively. Our work demonstrates that assumptions about iWUE in Earth system models significantly affect photosynthesis and its response to climate. Further, the response of photosynthesis to iWUE depends on which components of the model are included, therefore studies of iWUE impacts on historical or future photosynthesis can not be generalized across model configurations.

Atmospheric rivers (ARs) transport vast amounts of moisture from low to high latitude regions. One region particularly impacted by ARs is Interior Alaska (AK). We analyze the impact of ARs on the annual river ice breakup date for 25 locations in AK. We investigate the AR-driven rise in local air temperatures and explore the relationship between ARs and precipitation, including extremes and interannual variability. We found that AR events lead to an increase in local air temperatures for up to one week (by ≈ 1°C). Interannually, ARs account for 36% of total precipitation, explain 48% of precipitation variability, and make up 57% of extreme precipitation events. By estimating the heat transfer between winter precipitation and the river ice surface, we conclude that increased precipitation during the coldest period of the year delays river ice breakup dates, while precipitation occurring close to the breakup date has little impact on breakup timing.

As high-resolution geospatiotemporal data sets from observatory networks, remote sensing platforms, and computational Earth systems increase in abundance, fidelity, and richness, machine learning approaches that can fully utilize increasingly powerful parallel computing resources are becoming essential for analysis and exploration of such data sets. We explore one such approach, applying a state-of-the-art distributed memory parallel implementation of Support Vector Machine (SVM) classification to large remote-sensing data sets. We have used MODIS 8-day surface reflectance (MOD09A1) and land surface temperature (MOD11A2) for classifying wildfires over Alaska and California. Monitoring Trends in Burn Severity (MTBS) burn perimeter data was used to set boundaries of burned and unburned areas for our two-class problem. MTBS covers years from 1984-2019, recording only fires over 1000 acres or greater in the western United States. We seek to find a parallel computing solution (using the PermonSVM solver, described below) to accurately classify wildfires and find smaller unrecorded wildfires. An initial assessment for wildfire classification over interior Alaska shows that PermonSVM has an accuracy of 96% and over 5000 false positives (i.e., fires unrecorded in MTBS). Next steps include mapping larger regions over Alaska and California and understanding the tradeoffs of scalability and accuracy. The parallel tool we employ is PermonSVM, which is built on top of the widely-used open source toolkit PETSc, the Portable, Extensible Toolkit for Scientific Computation. Recent developments in PETSc have focused on supporting cutting-edge GPU-based high-performance computing (HPC) architectures, and these can be easily leveraged in PermonSVM by using appropriate GPU-enabled matrix and vector types in PETSc. We achieve significant GPU speedup for the SVM calculations on the Summit supercomputer at Oak Ridge National Laboratory – currently one of the best available “at scale” proxies for upcoming exascale-class supercomputers – and are actively working to further improve computational efficiency on Summit as well as on prototype exascale node architectures.