Using data collected during the

2002 International H2O Project, this research focused on moisture transport

from grassland and cropland environments to the atmosphere. Land surface and

numerical weather prediction models often represent the moisture exchange from

vegetated surfaces using the Jarvis scheme with a parameterized minimum canopy

resistance (r_{c min}).

While the parameterized values are

held constant in the models, the results of this study suggests that r_{c min} is

highly variable both spatial and temporally at diurnal and longer timescales. Analyses

of r_{c min} derived from the observational data also

suggests that the variability results from both changing environmental

conditions - some of which, like soil moisture content, are considered by the

Jarvis scheme – and changing vegetation characteristics. Furthermore, a spatiotemporal

average across the entire research domain suggests that the parameterized value

(40 s m^-1) used in the Noah land surface model (Noah_LSM) is

significantly lower than the

r_{c min} derived

from observations (100 s m^-1). Additional analyses conducted using

the Noah_LSM show that the parameterization of r_{c min} has

a substantial impact on the modeled latent heat flux and the partition of the

surface energy budget. Based on these results, methods of improving land

surface models, such as the Noah_LSM, that estimate transpiration using Jarvis-type

relationships include incorporating a time varying value for r_{c min} or

replacing the Jarvis scheme with a physiologically-based estimate of canopy

resistance.