Mountain Temperatures at Fine Spatial Scales
Objectives: Running a standard weather station in the high mountains is very costly, and as a result, few measurements exist in mountainous regions. However, innovative new technology, deployed on existing structures (trees) instead of towers, may provide a low-cost key to filling in the gaps. First, we will test the stability, accuracy, and precision of various low-cost sensors against a standard instrument. Second, we will test several different radiation shield designs. Finally, we will deploy these temperature sensors in a variety of elevations and topographic settings in both the Niwot Ridge long-term ecological research (LTER) station and the Fraser Experimental Forest. We will then work across different scales to learn how localized temperature patterns on both sides of the continental divide relate to large-scale weather patterns, and how well long-term meteorological stations represent the surrounding topography.
Background and importance
In the western United States, over half of the water supply is derived from mountain snowmelt, where the snow delays runoff and provides water in the spring and summer, when it is needed most. In recent decades, snowmelt runoff timing has advanced one to three weeks earlier in mountainous catchments across western North America, responding primarily to temperature changes. However, these general trends are quite variable between basins, and at high elevations, little is known about the spatial and temporal variations of critical processes like snowmelt, runoff, or even temperature.
Mountains are spatially complex and sparsely sampled. Temperatures are usually interpolated from distant stations, assuming a constant decrease in air temperature of 6.5OC per 1000 m elevation gain. However, examination of observed surface temperatures indicates that they differ diurnally, synoptically, and seasonally and do not always increase linearly with elevation, which can have profound impacts on snowmelt and runoff forecasts. For example, Singh (1991) found that changes of 1OC per km in the lapse rate for the Beas watershed (345 km2, spanning 1900 to 5400 m elevation) produced variations of 28-37% in the modeled snowmelt runoff over a two-month period.
Due to cost restrictions, prior studies have focused on long-term temperature changes at one measurement site or spatially-extensive measurements for a period of less than a week. To be useful, studies must span both spatial and temporal scales. Fortunately, innovative new instruments, such as the Onset Tidbit and the iButton, are inexpensive, self-recording, and can be densely deployed across mountain landscapes for extended time periods. A collection of temperature sensors deployed in Yosemite National Park, California for the past two winters have shown spatial patterns in average daily temperature that differ from the standard lapse rate by over 5?C, with the most marked differences occurring during periods with inversions. Temperature variations such as these are controlled primarily by local topography, vegetation, and larger-scale weather patterns. Thus, we hope to establish relationships between these patterns, local temperatures, and temperature measured at long-term meteorological stations. We can then use these relationships to better interpret data at long-term stations, to better interpolate temperature across the landscape.
The strength of these new sensors is in their ability to enhance the interpretation of existing long-term, real-time weather stations. Thus, we propose to deploy 50 sensors each in two established study areas: the LTER station at Niwot Ridge, on the east side of the continental divide, which has 4 climate stations that have been operational since 1952, and the Fraser Experimental Forest on the west side of the divide, which was the site of extensive field measurements during the Cold Land Processes Experiment and has 8 weather stations operational since 1969. Together, the two study sites will allow us to compare processes on the west and east sides of the Rocky Mountains. Instruments will be suspended from tree branches, using dental floss and upside-down funnels for radiation shields. Different configurations of attachments and shields will be tested, including an examination of how trees influence the measurements (by deploying instruments on existing meteorological towers and on nearby trees). Each site will be documented with GPS measurements and photographs. Sites will sample a wide-variety of elevations and topographic characteristics. Scientists will return to the sites for repeat photo-documentation and to download instrument data at approximately 4-month intervals. Large-scale weather patterns will be determined using the NCEP-NCAR reanalysis data,. Local winds, radiative energy balance, and relative humidities will be determined from the existing weather towers. Spatial patterns will be analyzed within a statistical GIS framework, and dense temperature measurements will be compared with those estimated by Glen Listonís MicroMet model for distributing temperatures in complex terrain.
Expected outcome and impact: We expect both the Hobo and iButton sensors to perform well (based on preliminary tests by Jason Hubbart at the University of Idaho). We expect to observe repeatable spatial patterns in temperature that will depend on local topography and winds, which will vary between the west and east sides of the continental divide. The immediate outcome will be improved ability to distribute temperature spatially, which can be used to improve snowmelt forecast models for the streams of the Fraser and Niwot Ridge areas, both of which provide water for the Boulder/Denver area. However, the impacts of this study are expected to be much wider. Once the value of these small sensors is demonstrated, they can be deployed in any watershed and will fundamentally change the way mountain observing systems are designed.
References: Singh, P. 1991: A temperature lapse rate study in Western Himalayas. Hydrology, Jour. of Indian Assoc. of Hydrologists, 14, 156-163.