The sensitivity of modeled snow accumulation and melt to precipitation phase methods across a climatic gradient
<p>A critical component of hydrologic modeling in cold and temperate regions is partitioning precipitation into snow and rain, yet little is known about how uncertainty in precipitation phase propagates into variability in simulated snow accumulation and melt. Given the wide variety of methods...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2019-09-01
|
| Series: | Hydrology and Earth System Sciences |
| Online Access: | https://www.hydrol-earth-syst-sci.net/23/3765/2019/hess-23-3765-2019.pdf |
| Summary: | <p>A critical component of hydrologic modeling in cold and
temperate regions is partitioning precipitation into snow and rain, yet
little is known about how uncertainty in precipitation phase propagates into
variability in simulated snow accumulation and melt. Given the wide variety
of methods for distinguishing between snow and rain, it is imperative to
evaluate the sensitivity of snowpack model output to precipitation phase
determination methods, especially considering the potential of snow-to-rain
shifts associated with climate warming to fundamentally change the hydrology
of snow-dominated areas. To address these needs we quantified the
sensitivity of simulated snow accumulation and melt to rain–snow
partitioning methods at sites in the western United States using the
SNOWPACK model without the canopy module activated. The methods in this
study included different permutations of air, wet bulb and dew point
temperature thresholds, air temperature ranges, and binary logistic
regression models. Compared to observations of snow depth and snow water equivalent (SWE), the
binary logistic regression models produced the lowest mean biases, while
high and low air temperature thresholds tended to overpredict and
underpredict snow accumulation, respectively. Relative differences between
the minimum and maximum annual snowfall fractions predicted by the different
methods sometimes exceeded 100 % at elevations less than 2000 m in the
Oregon Cascades and California's Sierra Nevada. This led to ranges
in annual peak SWE typically greater than 200 mm,
exceeding 400 mm in certain years. At the warmer sites, ranges in snowmelt
timing predicted by the different methods were generally larger than 2 weeks, while ranges in snow cover duration approached 1 month and greater.
Conversely, the three coldest sites in this work were relatively insensitive
to the choice of a precipitation phase method, with average ranges in annual
snowfall fraction, peak SWE, snowmelt timing, and snow cover duration of less
than 18 %, 62 mm, 10 d, and 15 d, respectively. Average ranges in snowmelt
rate were typically less than 4 mm d<span class="inline-formula"><sup>−1</sup></span> and exhibited a small
relationship to seasonal climate. Overall, sites with a greater proportion
of precipitation falling at air temperatures between 0 and
4 <span class="inline-formula"><sup>∘</sup></span>C exhibited the greatest sensitivity to method selection,
suggesting that the identification and use of an optimal precipitation phase
method is most important at the warmer fringes of the seasonal snow zone.</p> |
|---|---|
| ISSN: | 1027-5606 1607-7938 |