IHOP 2002 was conducted over the southern Great-Plains region to study water vapor distribution in space and time to improve quantitative precipitation forecasting to improve flash-flood warnings. Three main transects were the primary focus of the study. One was in northern Texas and the western Oklahoma panhandle, the second was located in centeral Kasas and the third in eastern Kasas with increasing precipitation from west to east. Each transect had micrometeorologic surface flux stations on several land-use types and were the target of aircraft missions to characterize many aspects of water vapor distibution in the atmosphere. Land-surface characterization was also performed, including pole-mounted multi-spectral radiometer measurements from which NDVI was calculated. Ten NDVI samples were taken during clear skies along consistent transects at flux stations for which averages are presented in figures 1-3 below. Land-surface characterization was performed by Joe Alfieri, Joshua Uebelherr and Heather McEntire; data post processing was completed by Joe Alfieri.
Note the dramatic increase in NDVI from west to east along the precipitation gradient for figures 1-3 and the decrease in NDVI for ripening winter wheat crops in figure 2.
Fig. 1: Low NDVI values can be seen for bare soil (1-W) early on with some vegetation sprouting after rain in mid-July. Site 2-W was a Conservation Recovery Site and was predominantly vine-weed with substantial bare-soil. Site 3-W was a grazed sage/grass site with some yucca and cactus and sandy soil. The heavily grazed pasture (10-W) was mostly dead clods of grass prior to early July when new grass began to grow following precipitation.
Fig. 2: Note the counter trends between the senescent winter wheat and the green-up of the C4 dominated sparse grassland.
Fig. 3: Cloud cover prevented NDVI from being measured in May, though high NDVI values wer recorded for all eastern Kansas grassland sites.
CASES-97
The Cooperative Atmospheric Exchange Study 1997 (CASES-97) investigated land-atmophere interaction in the Walnut Creek watershed in Butler County, Kansas. The project was carried out by the National Center for Atmospheric Research over several land use types. Below are figures I have prepared from the grassland site 8 from CASES-97.
Fig. 1: Bulk canopy conductance (gc) was calculated from a rearranged form of the Penman-Montieth model for dry canopy conditions when net energy exceeded 50 W/m2. Note the response of gc to precipitation cycles and associated changes in soil moisture as well as an apparent decoupling of gc in late May when vegetation growth increases.
Fig. 5: Relationship of NOAA 14 satellite AVHRR NDVI with gc for days when data was available. Despite the complex behavior of gc seen in fig. 4, NDVI and gc have a good relationship, although there are few points for comparison.
