9.2 Using high-resolution land data assimilation system to improve prediction of soil temperature and moisture for agricultural applications (2008

9.2

Using high-resolution land data assimilation system to improve prediction of soil temperature and moisture for agricultural applications

Ying Zhang, NCAR, Boulder, CO; and F. Chen, B. Myers, K. W. Manning, and M. Barlage

The NCAR High Resolution Land Data Assimilation System (HRLDAS) is further enhanced

to provide the forecast of soil temperature and soil moisture as input to agricultural decision

support systems (pest control, seeding, etc). Various observations and high-resolution

land-use and soil texture fields were used to drive HRLDAS

on 4-km domain in CONUS for a retrospective period (2005-2006) and for

realtime forecast. The hindcast of soil temperature and moisture from March to August 2006

was evaluated against the soil n observations from the Soil Climate Analysis Network (SCAN), which

has more than 116 stations located in 39 states and collects hourly atmospheric, soil

temperature and soil moisture data.

Evaluation statistics reveal that HRLDAS is able to capture both the observed diurnal cycle

and long-term evaluation of soil temperature and its vertical structures. Among atmospheric forcing

conditions used to drive HRLDAS, the surface air temperature, hourly precipitation and solar

radiation play important roles in determining the evolution of soil temperature and moisture. Also, HRLDAS

shows significant sensitivity to the specification of thermal bottom-boundary conditions, the number of

vertical layers in HRLDAS, and vegetation phenology. We will discuss a number of efforts to improve

the forecast of soil temperature and soil moisture, which include the use of 1) MODIS vegetation

products (vegetation cover and leaf area index), 2) high-resolution climatology air temperature

for specifying bottom soil temperature, and 3) Kalman filter technique for assimilating soil moisture

and soil temperature.