Representing atmospheric moisture content in the mountains: Examination using distributed sensors in the Sierra Nevada, California

Abstract

Atmospheric moisture content is a critical factor in both the water balance and the energy balance for a river basin. Despite its importance to hydrology, atmospheric moisture is sparsely measured, particularly in the mountains. Since few observations exist, numerous empirical methods have been developed to estimate relative humidity (RH) or the dewpoint temperature. However, most of these algorithms were developed in continental regions and may have limited accuracy outside the region where they were developed. Furthermore, future changes in atmospheric moisture content may reduce our ability to rely on empirically determined relationships. Alternative options include installing more in situ sensors, looking at nearby free air measurements, and/or running a numerical weather model. We compared densely-distributed measurements of dewpoint temperatures in two study sites over three years in a semi-arid, maritime mountain range (Sierra Nevada, California) against: (1) simple empirical algorithms, (2) the Parameter-elevation Regressions on Independent Slopes Model (PRISM) linear regression data sets based on observational data, (3) the Weather Research and Forecasting (WRF) mesoscale model, and (4) radiosonde data. Empirical algorithms that used only one sea-level measurement of dewpoint to extrapolate to higher elevations, on average overestimated moisture in the basin, displaying median biases of daily dewpoint temperatures up to 10.5°C. These algorithms were subject to errors both from misrepresenting the linear rate of moisture loss with elevation and, on some days, from assuming the dewpoint temperature followed a linear pattern at all. These methods used assumptions that were empirically-derived in other climates. PRISM improved upon these methods by using local observations to determine the local average lapse rate, with median bias values of -0.3°C and 2.2°C in our study sites. Empirical algorithms that derived dewpoint from air temperature showed a significant seasonal variation in performance. Assuming uniform advection of moisture from the Pacific does not capture the moisture dynamics in the Sierra Nevada. Radiosonde readings showed large biases from observations, and a wide range of day to day error. WRF improved on the free-air data, performing well in representing both the overall trends in the basin (with median biases of -0.9°C and -1.0°C in our study sites) and displaying the smallest range of error throughout the year. The impact of errors in dewpoint temperature estimation on hydrology is by applying the Distributed Hydrology Soils and Vegetation Model (DHSVM) to the upper Tuolumne River Watershed in Yosemite. A ±2°C bias in dewpoint temperatures resulted in an average ±3 day shift in snowmelt timing and change annual streamflow volumes by ±1%. When modeling a geographically simple basin, one base station within the basin paired with PRISM lapse rates will be representative of overall moisture trends most of the time, biased by -0.3°C in one study site. However, if the basin is more geographically complex, with air masses varying due to predominant weather patterns, micro-topography, and transport along the mountain range, a physically-resolved model such as WRF is necessary to represent dewpoint variations. To reduce average modeled bias in a basin, the simplest method is to add a high-elevation station that records dewpoint temperatures and use that to model the average dew point temperature decline with elevation.