Tuesday, August 4, 2009

PS 38-160: Use of a digital camera to estimate heterogeneous temperatures and heat flux in a forest understory soil

Eric A. Graham, Eric M. Yuen, and Yeung Lam. University of California, Los Angeles

Background/Question/Methods

Soil temperature drives microbial activity, plant germination, rooting, and CO2 fluxes.  Soil energy balance is also used in calculations of eddy flux measurements.  Because soil surface temperatures can be spatial heterogeneous due to uneven heating from partial shading by an overstory, calculating soil energy balances over large areas and predicting sub-surface temperatures in a forested environment is difficult.

Using the analytical models for homogeneously conductive material introduces errors when trying to accurately predict sub-surface temperatures from those at the surface.  However, complex patterns of soil surface and sub-surface temperatures due to partial shading by an overstory can be modeled using after the decomposition of the signals into a series of simpler sine waves of different frequencies.  Simultaneously, the use of a digital camera to detect areas of solar radiation, coupled with measurements of air temperature, can be used to predict soil surface temperature.  After soil surface and subsurface temperatures are calculated, large areas estimates of energy balance can then be made.

Results/Conclusions

A surface scan of temperatures along a 12 m the transect was made using a robotic platform with energy balance sensors and a downwardly facing infrared thermometer.  Belowground temperature and heat flux was also measured at four locations along the transect.  Data were captured continuously for 36 hours by a set of networked dataloggers simultaneously with images captured by a digital camera of the understory.  Periodic but irregular fluctuations in soil surface temperatures occurred in the forested environment primarily as the result of daily solar input.  A modified temperature model was used to successfully predict deeper-layer temperatures using different numbers of component sine waves derived from the Fourier transforms of the temperature signal for overlying soil layers.  A large spatial extent of soil energy balance was then modeled using the time series of images captured of the understory to determine the extent of solar radiation.