Nicholas E. Wayand

Intercomparison of Meteorological Forcing Data from Empirical and Mesoscale Model Sources in the North Fork American River Basin in Northern Sierra Nevada, California*

AbstractThe data required to drive distributed hydrological models are significantly limited within mountainous terrain because of a scarcity of observations. This study evaluated three common configurations of forcing data: 1) one low-elevation station, combined with empirical techniques; 2) gridded output from the Weather Research and Forecasting Model (WRF); and 3) a combination of the two. Each configuration was evaluated within the heavily instrumented North Fork American River basin in California during October–June 2000–10. Simulations of streamflow and snowpack using the Distributed Hydrology Soil and Vegetation Model (DHSVM) highlighted precipitation and radiation as variables whose sources resulted in significant differences. The best source of precipitation data varied between years. On average, the WRF performed as well as the single station distributed using the Parameter Regression on Independent Slopes Model (PRISM). The average percent biases in simulated streamflow were 3% and 1%, for con...

Diagnosis of insidious data disasters

Everyone taking field observations has a story of data collection gone wrong, and in most cases, the errors in the data are immediately obvious. A more challenging problem occurs when the errors are insidious, i.e., not readily detectable, and the error-laden data appear useful for model testing and development. We present two case studies, one related to the water balance in the snow-fed Tuolumne River, Sierra Nevada, California, combined with modeling using the Distributed Hydrology Soil Vegetation Model (DHSVM); and one related to the energy balance at Snoqualmie Pass, Washington, combined with modeling using the Structure for Unifying Multiple Modeling Alternatives (SUMMA). In the Tuolumne, modeled streamflow in 1 year was more than twice as large as observed; at Snoqualmie, modeled nighttime surface temperatures were biased by about +10°C. Both appeared to be modeling failures, until detective work uncovered observational errors. We conclude with a discussion of what these cases teach us about science in an age of specialized research, when one person collects data, a separate person conducts model simulations, and a computer is charged with data quality assurance.

A meteorological and snow observational data set from Snoqualmie Pass (921 m), Washington Cascades, USA

We introduce a quality controlled observational atmospheric, snow, and soil data set from Snoqualmie Pass, Washington, USA, to enable testing of hydrometeorological and snow process representations within a rain-snow transitional climate where existing observations are sparse and limited. Continuous meteorological forcing (including air temperature, total precipitation, wind speed, specific humidity, air pressure, and short and longwave irradiance) are provided at hourly intervals for a 24 year historical period (water years 1989–2012) and at half-hourly intervals for a more recent period (water years 2013–2015), separated based on the availability of observations. The majority of missing data were filled with biased-corrected reanalysis model values (using NLDAS). Additional observations include 40 years of snow board new snow accumulation, multiple measurements of total snow depth, and manual snow pits, while more recent years include subdaily surface temperature, snowpack drainage, soil moisture and temperature profiles, and eddy covariance-derived turbulent heat flux. This data set is ideal for testing hypotheses about energy balance, soil, and snow processes in the rain-snow transition zone.

Modeling the influence of hypsometry, vegetation, and storm energy on snowmelt contributions to basins during rain‐on‐snow floods

Point observations and previous basin modeling efforts have suggested that snowmelt may be a significant input of water for runoff during extreme rain-on-snow floods within Western U.S. basins. Quantifying snowmelt input over entire basins is difficult given sparse observations of snowmelt. In order to provide a range of snowmelt contributions for water managers, a physically-based snow model coupled with an idealized basin representation was evaluated in point simulations and used to quantify the maximum basin-wide input from snowmelt volume during flood events. Maximum snowmelt basin contributions and uncertainty ranges were estimated as 29% (11-47%), 29% (8-37%), and 7% (2-24%) of total rain plus snowmelt input, within the Snoqualmie, East North Fork Feather, and Upper San Joaquin basins, respectively, during historic flooding events between 1980 and 2008. The idealized basin representation revealed that both hypsometry and forest cover of a basin had similar magnitude of impacts on the basin-wide snowmelt totals. However, the characteristics of a given storm (antecedent SWE and available energy for melt) controlled how much hypsometry and forest cover impacted basin-wide snowmelt. These results indicate that for watershed managers, flood forecasting efforts should prioritize rainfall prediction first, but cannot neglect snowmelt contributions in some cases. Efforts to reduce the uncertainty in the above snowmelt simulations should focus on improving the meteorological forcing data (especially air temperature and wind speed) in complex terrain. This article is protected by copyright. All rights reserved.

Improving simulations of precipitation phase and snowpack at a site subject to cold air intrusions: Snoqualmie Pass, WA

Low-level cold air from eastern Washington often flows westward through mountain passes in the Washington Cascades, creating localized inversions and locally reducing climatological temperatures. The persistence of this inversion during a frontal passage can result in complex patterns of snow and rain that are difficult to predict. Yet, these predictions are critical to support highway avalanche control, ski resort operations, and modeling of headwater snowpack storage. In this study we used observations of precipitation phase from a disdrometer and snow depth sensors across Snoqualmie Pass, WA, to evaluate surface-air-temperature-based and mesoscale-model-based predictions of precipitation phase during the anomalously warm 2014-2015 winter. Correlations of phase between surface-based methods and observations were greatly improved (r2 from 0.45 to 0.66) and frozen precipitation biases reduced (+36% to -6% of accumulated snow water equivalent) by using air temperature from a nearby higher-elevation station, which was less impacted by low-level inversions. Alternatively, we found a hybrid method that combines surface-based predictions with output from the Weather Research and Forecasting mesoscale model to have improved skill (r2 = 0.61) over both parent models (r2 = 0.42 and 0.55). These results suggest that prediction of precipitation phase in mountain passes can be improved by incorporating observations or models from above the surface layer.

Diagnosing Snow Accumulation Errors in a Rain‐Snow Transitional Environment with Snow Board Observations

Diagnosing the source of errors in snow models requires intensive observations, a flexible model framework to test competing hypotheses, and a methodology to systematically test the dominant snow processes. We present a novel process-based approach to diagnose model errors through an example that focuses on snow accumulation processes (precipitation partitioning, new snow density, and snow compaction). Twelve years of meteorological and snow board measurements were used to identify the main source of model error on each snow accumulation day. Results show that modeled values of new snow density were outside observational uncertainties in 52% of days available for evaluation, while precipitation partitioning and compaction were in error 45% and 16% of the time, respectively. Precipitation partitioning errors mattered more for total winter accumulation during the anomalously warm winter of 2014-2015, when a higher fraction of precipitation fell within the temperature range where partition methods had the largest error. These results demonstrate how isolating individual model processes can identify the primary source(s) of model error, which helps prioritize future research.