Geoscience Reference
In-Depth Information
Figure 24.8
Cumulative observed average precipitation for the region 30-40°N 92.5-107.5°W compared with ensemble-
average cumulative precipitation calculated by the WRF model with the NOAA land surface scheme (labeled 'control'),
a NOAA scheme including simulation of dynamic vegetation and a NOAA scheme including simulation of both dynamic
vegetation and ground water interactions. (Adapted from Jiang
et al
., 2009.)
Among these four approaches (a) arguably has least accuracy in representing the
dynamic movement of the groundwater table, and the two-way coupling in (c) and
(d) is likely preferable. Of these last two, (d) is favored on the grounds of
computational efficiency unless the groundwater table is close to the bedrock.
Despite the limitations involved in all these approaches, these several studies have
together suggested the importance and potential impact of interactions between
groundwater and surface water on surface water and energy budgets, surface runoff
generation, soil moisture, groundwater recharge and subsurface flow characteristics
at different spatial scales (e.g., Jiang
et al.
, 2009), see Fig. 24.8. However, work in
this area is ongoing and the most effective way forward is yet to be defined.
Hydraulic redistribution (or hydraulic lift) is the passive movement of water from
roots into soil layers or from soil layers into roots. It can occur upwards, downwards
or laterally depending on the water potential gradients in the soil (e.g., Burgess
et al.
,
1998; Hultine
et al.
, 2003a; 2003b; Leffler
et al.
, 2005). It is now believed that
hydraulic redistribution may have significant impact on weather and climate
through its impact on evapotranspiration. Hasler and Avissar (2007) showed, for
example, that global and regional meteorological models both currently tend to
overestimate the dry season water stress in the Amazon basin relative to eddy
covariance flux measurements from eight towers, probably due to their
misrepresentations of the soil water and plant processes. While Lee
et al.
(2005)
incorporated a simple empirical formulation of hydraulic redistribution into version
2 of the NCAR Community Climate Model coupled to the Community Land Model
