 |
| Titel |
Explicit representation of groundwater process in a global-scale land surface model to improve the prediction of water resources |
| VerfasserIn |
Sujan Koirala, Pat J.-F. Yeh, Shinjiro Kanae, Taikan Oki |
| Konferenz |
EGU General Assembly 2010
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 12 (2010) |
| Datensatznummer |
250041063
|
|
|
|
|
|
| Zusammenfassung |
| Groundwater is the most important freshwater resource and its relevance can be viewed in
two aspects; a pervasive and seemingly abundant storage of freshwater and a consistent
source of surface water in dry season in form of base runoff, which is nothing but a
groundwater reservoir discharging into rivers. For proper estimation of groundwater
resources in current and future climate conditions, explicit physically-based representation of
groundwater process is necessary in models. Traditionally, global land surface models
(LSMs) mainly focused on energy balance at land surface and often they have simplified
runoff scheme largely neglecting the groundwater process. In this study, an explicit shallow
groundwater representation was integrated into a LSM, Minimal Advanced Treatments of
Surface Interaction and Runoff (MATSIRO).
The coupled model was applied in global-scale to evaluate the role and relevance of
explicit groundwater representation on the prediction of groundwater-based freshwater
resources. Three aspects of concern were seasonal variation of discharge, simulation of low
flow, and estimation of net groundwater recharge. The results were compared with a
commonly used approach in land surface modeling, in which gravity drainage from
unsaturated soil column is assumed to be base runoff. The results showed that groundwater
representation significantly improved the partitioning of runoff into surface and sub-surface
component, and subsequently seasonal variation of discharge in majority of 20 major
river basins of the world. In the prediction of monthly river discharge, the model
with explicit groundwater representation out-performed the model with gravity
drainage in 18 out of 20 target river basins. The most significant improvements
were in river basins of semi-arid regions (e.g., Darling, Orange, and Zambezi river
basins), and the river basins with marked dry season (e.g., Ganges and Mekong river
basins) where upward capillary flow from groundwater reservoir to unsaturated
soil column is substantial. Under such conditions, the gravity drainage assumption
over-predicted net drainage flux from unsaturated zone to groundwater reservoir and
consequently base runoff was over-predicted. Furthermore, using MATSIRO with
groundwater representation, direct groundwater runoff from coastal region to ocean was
estimated to be 5788 km3-year which is around 15 % of estimated global total
runoff. Also, comparison of daily flow duration curves showed that the prediction of
low flow, with probability of exceedance -¥ 90 %, was also enhanced with explicit
representation of groundwater process. Finally, global groundwater recharge was predicted
to be 31789 km3-year. If the water table is in equilibrium condition, long-term
mean groundwater recharge should be of similar magnitude to long-term mean base
runoff. Multi-model ensemble mean base runoff from second phase of Global Soil
Wetness Project (GSWP-2) is 30200 km3-year. Due to lack of validation data for
groundwater recharge on global scale, the estimation close to GSWP-2 multi-model
ensemble mean base runoff can be assumed to provide first order approximation.
Region-wise, humid regions (e.g., Amazon and Congo river basins) have the largest net
groundwater recharge. The net groundwater recharge is small for arid and semi-arid
regions mainly because of small precipitation input, high evaporative loss, and
strong upward flux from saturated zone (groundwater reservoir) to unsaturated
zone.
With enhanced prediction of monthly river discharge, low flow in daily temporal scale,
and net groundwater recharge, the MATSIRO land surface model with explicit representation
of groundwater process can be used in global-scale water resources assessment under current
and future climate conditions. |
|
|
|
|
|
|