Parameter Measurement and Estimation at Variable Scales: Example of Soil Temperature in Complex Terrain

Abstract:

The issue of matching measurement scale to application scale is long standing and frequently revisited with advances in instrumentation and computing power. In the past we have emphasized the importance of understanding the dominant processes and amount and nature of parameter variability when addressing these issues. Landscape-scale distribution of carbon and carbon fluxes is a primary focus of the Reynolds Creek Critical Zone Observatory (RC CZO). Soil temperature (T_s) is a critical parameter of generally unknown variability. Estimates of T_s are often based on air temperature (T_a), but it is understood that other factors control T_s, especially in complex terrain, where solar radiation may be a major driver. Data were collected at the Reynolds Creek Experimental Watershed (RCEW), which is 240 km² in extent and covers a 1000 m elevation range. We used spatially extensive T_s data to evaluate correlations with T_a (915 m elevation gradient) on aspect neutral sites with similar vegetative cover. Effects of complex terrain were evaluated using a combination of fixed point measurements, fiber optic distributed temperature sensing and periodic, spatially distributed point measurements. We found that T_s over the elevation gradient followed T_a closely. However, within small subwatersheds with uniform T_a, T_s may be extremely variable, with a standard deviation of 8° C. This was strongly related to topographically associated land surface units (LSU’s) and highly seasonal. Within LSU variability was generally low while there were seasonally significant differences between LSU’s. The mean annual soil temperature difference between LSU’s was greater than that associated with the 915 m elevation gradient. The seasonality of T_s variability was not directly related to solar radiation effects but rather to variations in cover. Scaling T_s requires high resolution accounting of topography in this environment. Spatial patterns of soil carbon at the RCEW are consistent with this.