 |
| Titel |
Combined use of local and global hydrometeorological data with regional and global hydrological models in the Magdalena - Cauca river basin, Colombia |
| VerfasserIn |
Erasmo Rodriguez, Ines Sanchez, Nicolas Duque, Patricia Lopez, Alexander Kaune, Micha Werner, Pedro Arboleda |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250146450
|
| Publikation (Nr.) |
 EGU/EGU2017-10477.pdf |
|
|
|
|
|
| Zusammenfassung |
| The Magdalena Cauca Macrobasin (MCMB) in Colombia, with an area of about 257,000
km2, is the largest and most important water resources system in the country. With almost
80% of the Colombian population (46 million people) settled in the basin, it is the main
source of water for demands including human consumption, agriculture, hydropower
generation, industrial activities and ecosystems. Despite its importance, the basin has
witnessed enormous changes in land-cover and extensive deforestation during the last
three decades. To make things more complicated, the MCMB currently lacks a set
of tools to support planning and decision making processes at scale of the whole
watershed.
Considering this, the MCMB has been selected as one of the six different regional case
studies in the eartH2Observe research project, in which hydrological and meteorological
reanalysis products are being validated for the period 1980-2012. The combined use of the
hydrological and meteorological reanalysis data, with local hydrometeorological data
(precipitation, temperature and streamflow) provided by the National Hydrometeorological
Agency (IDEAM), has given us the opportunity to implement and test three hydrological
models (VIC, WFLOW and a Water Balance Model based on the Budyko framework) at the
basin scale. Additionally, results from the global models in the eartH2Observe hydrological
reanalysis have been used to evaluate their performance against the observed streamflow data.
This paper discusses the comparison between streamflow observations and simulations from
the global hydrological models forced with the WFDEI data, and regional models forced with
a combination of observed and meteorological reanalysis data, in the whole domain of the
MCMB.
For the three regional models analysed results show good performances for some
sub-basins and poor performances for others. This can be due to the smoothing of the
precipitation fields, interpolated from point daily rainfall data, the effect of horizontal
precipitation (not included in the models) and weaknesses in the models structures; for
example the poor performance of the VIC model in base flow dominated basins. In order to
improve these simulations a strategy based on a hydrological model ensemble is currently
being developed in the case study.
Results from the global models, show that these consistently tend to overestimate runoff.
This may be due to the coarse resolution used (50 km), biases in the ERA-Interim
precipitation forcing, and the different partitioning within the global models of the
precipitation into evapotranspiration and runoff. It is expected that within the Tier II
hydrological reanalysis, where the models will produce outputs at 25 km resolution, some
improvements may be identified. |
|
|
|
|
|
|