Michael Barlage

Land surface models (LSMs) are widely used to generate snow records, forecasts, and projections that support water research and management. The complex sources of uncertainty in LSMs motivate evaluations to better understand bias characteristics and identify underlying causes that guide model development. This study evaluates snow water equivalent (SWE) simulations over the western United States (WUS) from the Noah‐MP LSM, a key land component in widely used research and operational modeling systems. Noah‐MP driven by different atmospheric forcing data sets generally underestimates SWE relative to SNOw TELemetry observations, with biases partly explained by scale mismatches between point‐based measurements and model pixels. Despite systematic underestimation, Noah‐MP captures observed interannual variability of peak and mean SWE anomalies very well ( r = 0.99) and shows lower but still significant accuracy in simulating long‐term regional SWE trends and the spatial patterns of SWE trends ( r = 0.60). Comparisons with gridded reference data sets confirm Noah‐MP's high skill in simulating the regional daily timing ( r > 0.99), interannual variability, and pixel‐scale daily SWE ( r = 0.71–0.81). However, the simulations underestimate SWE by 9%–16%, with peak SWE occurring ∼4 days earlier and snow disappearing ∼35 days earlier than observed, reflecting overly rapid ablation across both cold and warm seasons. Spatial clustering analyses reveal regional heterogeneity in SWE bias characteristics and sensitivities to LSM parameterizations and meteorological forcing uncertainties. The presence of regional overestimation suggests that applying uniform model adjustments may degrade performance in certain areas. An updated combination of Noah‐MP physics options shows accuracy improvements, reducing both accumulation and ablation errors.

Subseasonal‐to‐seasonal (S2S) forecasts are critically important for anticipating socioeconomic impacts but remain a significant scientific challenge. Reliable forecasts beyond a 2‐week timeframe require both accurate initial conditions and realistic representations of coupled Earth system processes. In particular, the Earth's land surface provides a key source of predictability for S2S predictions through land‐atmosphere interactions. Here, we present a 25‐year global 60‐km simulation using the Model for Prediction Across Scales (MPAS) coupled to the Noah‐MP land surface model, forced with ERA‐5 sea surface temperature and sea ice boundary conditions. This study evaluates the model's climatology, systematic biases, and ability to represent land‐atmosphere interactions, all of which are critical for future S2S applications. MPAS‐NoahMP captures well the annual surface temperature in the tropics and midlatitudes, while winter temperatures in North America and Asia are sensitive to sea ice thickness and snowpack thermal properties. The global precipitation pattern is also effectively reproduced by MPAS‐NoahMP, with slight dry biases in the tropics of the Eastern Pacific, Amazon, Congo Basin, and Maritime Continent, and wet biases in the tropical Indian ocean. The model accurately simulates spatial patterns of land‐atmosphere coupling indices and hotspots, including winter snow‐temperature coupling and summer soil moisture‐latent heat flux coupling. Furthermore, MPAS‐NoahMP reasonably represents the major modes of tropical variability, including the El Nino‐Southern Oscillation and the Madden‐Julian Oscillation. This study is the first to establish a long‐term MPAS‐NoahMP reference data set for understanding model biases and assessing land‐atmosphere interactions, providing a foundation for guiding future improvements to the coupled system for S2S applications.

Convection‐permitting modeling has become a cornerstone for improving representation of land‐atmosphere processes, predicting weather extremes, and creating long‐term hydrometeorological data sets. The convection‐permitting Weather Research and Forecasting (WRF) modeling with the NoahMP land‐surface scheme is widely used for kilometer‐scale hydroclimatic applications. We identify persistent large warm biases (up to ∼8°C) in daily minimum surface air temperature ( T _min ) in convection‐permitting WRF/NoahMP simulations across the western, central, and southeastern US. Motivated by the potential of soil organic matter (SOM) to mitigate nighttime warm bias, we implement a physical representation of this previously missing component and assess its effects on land‐surface fluxes and surface air temperature. Including SOM significantly changes surface soil thermal properties and reduces T _min warm bias (by 75%) by mitigating the overestimated diurnal amplitude of ground heat flux (by 29%). This offers a physics‐based pathway to improve future kilometer‐scale simulations.

Atmospheric models generally exhibit cold biases in 2‐m air temperature (T2) across mountainous regions during snow seasons. This study evaluates whether snow cover errors are the primary driver for the persistent T2 cold bias in coupled WRF/Noah‐MP simulations across the western U.S (WUS) through analyses of T2, ground snow cover fraction (SCF), snow water equivalent (SWE), and surface albedo. Baseline simulations revealed widespread cold biases in maximum daily T2 (T2 _max ), mainly during November–April when snowpack influences the terrestrial energy budget (mean bias = −1.05°C to −1.32°C), with higher elevation areas showing biases up to −5.31°C. Cold biases across the WUS are reduced, on average, by 17%–24% in the simulations using recently enhanced snow compaction and ground SCF parameterizations (hereinafter ARnew) that tend to reduce snow cover and surface albedo. However, notable cold biases still persist in the ARnew simulations, especially in high‐elevation areas, despite underestimates of ground SCF and SWE. Statistical and information theory analyses show very weak connections between T2 _max biases and biases in ground SCF and SWE. Further, an experiment that imposes unrealistically low ground SCF (mean bias = −75%) still exhibits cold biases, particularly in high‐elevation areas (mean bias for Z > 2,000 m = −1.46°C). These results indicate that snow cover errors can modestly contribute to cold biases but are not their primary cause. Therefore, resolving the cold bias issue will require addressing factors beyond snow cover, potentially including uncertainties in canopy radiative transfer, surface turbulence, and other atmospheric processes in complex terrain.

Subseasonal-to-seasonal (S2S) forecasts are critically important for anticipating socio-economic impacts but remain a significant scientific challenge. Reliable forecasts beyond a two-week timeframe require both accurate initial conditions and realistic representations of coupled Earth system processes. In particular, Earth’s land surface provides a key source of predictability for S2S predictions through land-atmosphere interactions. Here, we present a 25-year global 60-km simulation with the Model for Prediction Across Scales (MPAS) coupled with Noah-MP land surface model, with ERA-5 sea surface temperature and sea ice boundary conditions. This study evaluates the model’s climatology, systematic biases, and its ability to represent land-atmosphere interactions, all of which are necessary for future S2S applications. MPAS-NoahMP well captures annual surface temperature in the tropics and midlatitudes, while winter temperature in North America and Asia is sensitive to sea ice thickness and snowpack thermal properties. The global precipitation pattern is also well captured by MPAS-NoahMP, with slight dry biases in the tropics over the Eastern Pacific, Amazon, and Africa Congo, and wet biases in the Indian ocean and Maritime continent. The model accurately simulates spatial patterns of land-atmosphere coupling indices and hotspots, including winter snow-temperature coupling and summer soil moisture-latent heat flux coupling. Furthermore, MPAS-NoahMP reasonably captures the major modes of tropical variability, including the El Nino and Southern Oscillation and the Madden Julian Oscillation. This study is the first to establish a long-term MPAS-NoahMP reference dataset for understanding model biases and assessing land-atmosphere interactions, providing a foundation for extending the coupled MPAS-NoahMP system to future S2S applications.

Snow cover fraction (SCF) accuracy in land surface models (LSMs) impacts the accuracy of surface albedo and land‐atmosphere interactions. However, SCF is a large source of uncertainty, partially because of the scale‐dependent nature of snow depletion curves that is not parameterized by LSMs. Using the spatially and temporally complete observationally‐informed STC‐MODSCAG and Snow Data Assimilation System data sets, we develop a new scale‐aware ground SCF parameterization and implement it into the Noah‐MP LSM. The new scale‐aware parameterization significantly reduces ground SCF errors and the scale‐dependence of errors in the western U.S (WUS) compared with the baseline ground SCF formulation. Specifically, the baseline formulation overestimates ground SCF by 4%, 6%, 9%, and 12% at 1‐km, 3‐km, 13‐km, and 25‐km resolutions in the WUS, respectively, whereas biases from the enhanced scale‐aware scheme are reduced to 0%–2% in box model simulations and do not exhibit a relationship with spatial scales. Noah‐MP simulations using the scale‐aware parameterization have smaller mean (peak) ground SCF biases than the baseline simulation by 1%–2% (3%–5%), with spatiotemporal variability depending on land cover, topography, and snow depth. Noah‐MP simulations using the enhanced scale‐aware parameterization remove the baseline WUS surface albedo overestimates of 0.01–0.03 in the 1‐km to 25‐km resolution simulations, relative to Moderate Resolution Imaging Spectroradiometer retrievals. The Noah‐MP ground SCF and surface albedo improvements due to the scale‐aware parameterization are found across most land cover classifications and elevations, indicating the enhanced ground SCF scheme can improve simulated snowpack and surface energy budget accuracy across a variety of WUS landscapes.

Lake surface conditions are critical for representing lake‐atmosphere interactions in numerical weather prediction. The Community Land Model's 1‐D lake component (CLM‐lake) is part of NOAA's High‐Resolution Rapid Refresh (HRRR) 3‐km weather/earth‐system model, which assumes that virtually all the two thousand lakes represented in CONUS have distinct (for each lake) but spatially uniform depth. To test the sensitivity of CLM‐lake to bathymetry, we ran CLM‐lake as a stand‐alone model for all of 2019 with two bathymetry data sets for 23 selected lakes: the first had default (uniform within each lake) bathymetry while the second used a new, spatially varying bathymetry. We validated simulated lake surface temperature (LST) with both remote and in situ observations to evaluate the skill of both runs and also intercompared modeled ice cover and evaporation. Though model skill varied considerably from lake to lake, using the new bathymetry resulted in marginal improvement over the default. The more important finding is the influence bathymetry has on modeled LST (i.e., differences between model simulations) where lake‐wide LST deviated as much as 10°C between simulations and individual grid cells experienced even greater departures. This demonstrates the sensitivity of surface conditions in atmospheric models to lake bathymetry. The new bathymetry also improved lake depths over the (often too deep) previous value assumed for unknown‐depth lakes. These results have significant implications for numerical weather prediction, especially in regions near large lakes where lake surface conditions often influence the state of the atmosphere via thermal regulation and lake effect precipitation.

Most current land models approximate terrestrial hydrological processes as one‐dimensional vertical flow, neglecting lateral water movement from ridges to valleys. Such lateral flow is fundamental at catchment scales and becomes crucial for finer‐scale land models. To test the effect of incorporating lateral flow toward three‐dimensional representations of hydrological processes in the next generation land models, we integrate a water tracer model into the WRF‐Hydro framework to track water movement from precipitation to discharge and evapotranspiration. This hydrologic‐tracer integrated system allows us to identify the key mechanisms by which lateral flow affects the flow paths and transit times in WRF‐Hydro. By comparing modeling experiments with and without lateral routing in two contrasting catchments, we determine the impacts of lateral flow on the transit times of precipitation event‐water. Results show that with limited hydrologic connectivity, lateral flow extends the transit times by reducing (increasing) event‐water drainage loss (accumulation) in ridges (valleys) and allowing reinfiltration of infiltration‐excess flow, which is missing in most land models. On the contrary with high hydrologic connectivity, lateral flow can effectively accelerate the water release to streams and reduce the transit time. However, the transit times are substantially underestimated by the model compared with isotope‐derived estimates, indicating model limitations in representing flow paths and transit times. This study provides some insights on the fundamental differences in terrestrial hydrology simulated by land models with and without lateral flow representation.

The accuracy of snow density in land surface model (LSM) simulations impacts the accuracy of simulated terrestrial water and energy budgets. However, there has been little research that has focused on enhancing snow compaction in operationally used LSMs. A baseline snow simulation with the widely used Noah‐MP LSM systematically overestimates snow depth by 55 mm even after removing daily snow water equivalent (SWE) biases. To reduce uncertainties associated with snow compaction, we enhance the most sensitive Noah‐MP snow compaction parameter—the empirical parameter for compaction due to overburden (C _bd)—such that C _bd is calculated as a function of surface air temperature as opposed to a fixed value in the baseline simulation. This enhancement improves accuracy in simulated snow compaction across the majority of western U.S. (WUS) SNOTEL test sites (biases reduced at 88% of test sites), with modest bias reductions in cooler accumulation periods (biases reduced at 70% of test sites) and substantial improvements during warmer ablation periods (biases reduced at 99% of test sites). Relatively larger improvements during warm conditions are attributable to the default C _bd value being reasonable for cold temperatures (≤−5°C). Improvements in simulated snow depth and density with observations outside of the training sites and optimization periods support that the snow compaction enhancement is transferable in space and time. Differences between enhanced and baseline gridded simulations across the total WUS support that the enhancement can have important impacts on snowpack evolution, snow albedo feedback, and snow hydrology.