Evaluation of Water Balance Terms of the SEA

Evaluation of Water Balance Terms of the SEA

     Climatological forcing of the land surface (model) requires developing datasets that can be used to drive the models; preferably, with some understanding of the dynamics that produced those forcings. The question arises, how accurate at what resolutions are the data used to represent the forcings? Actual coupling can be passive (the provision of the necessary datasets) or active. To meet these requirements, we developed climate forcing data to drive the hydrology models. We collaborated with Dr. Ruby Leung, PNNL (as a sub-contract), to produce a data set from the Weather Research and Forecasting model (WRF), and then to use those results as part of a study to look at variations in precipitation across the region (Leung et al in prep). This collaboration also led to a data archive of model results at the UW, to be used in this and related studies, and it promoted the development of the direct coupling between WRF and VIC. Sonessa et al (in prep) used the WRF results in comparison to NCEP/NCAR reanalyses and ERA-Interim reanalysis to evaluate the water balance terms of the seven primary rivers of Southeast Asia. Beyenne et al (in prep) used the Mekong VIC forcings and climate projections from IPCC scenarios to examine the interaction of climate change and dams scenarios on the overall flow patterns of the Mekong, and changes in percent regulation.

Leung , R., et al. (in preparation). Diurnal and seasonal variations in precipitation and atmospheric heating in Asia based on observations and model simulations. In preparation. To be submitted to J. of Climatology. 

     Due to the influence of the large-scale circulation and its interactions with the complex terrain, precipitation over Asia varies at a wide range of temporal and spatial scales. The Weather Research and Forecasting model has been used to simulate the regional hydrological cycle in Asia at 18 km grid resolution using boundary conditions from a global analysis for 1997- 2008. Interior nudging was used to provide stronger constraints from the global reanalysis on the simulated large-scale circulation to better simulate seasonal and interannual variability associated with large-scale conditions. The diurnal precipitation variability in Asia has been analyzed using a combination of observations and the downscaled simulation to elucidate the dynamical and thermodynamical influence of orography and subsequent effects on the hydroclimate of the region. The TRMM precipitation data show distinct early morning rainfall timing at the foothills of the plateau and late afternoon/early evening timing on top of the plateau. Analysis of the simulation shows important influence of plateau-plain circulation and nocturnal drainage/upslope flow on the development of the boundary layer and diurnal rainfall. Comparison of the seasonal and interannual variability of rainfall among different remote sensing and ground based measurements and model simulation shows large differences and highlight the need to better constrain precipitation variability to yield improved estimates of the surface water budgets in the topographically diverse region.WRF and VIC Data Transfer and CouplingResults from the WRF modeling were transferred from PNNL to the UW.  The domain of focus covered the project region, 9.09 - 33.9285N and 91.1573 - 108.843 E (roughly 9 - 34 N and 91 - 109 E), for the period 01/1998 to 12/2006. Results were also catalogued for a broader domain, for future analysis.  Data includes: P, Tmax, Tmin, Wind Speed (vertical and horizontal components) data from 01/1998 to 12/2008.  The data is of 18km resolution. In preparation for more seamless operation, and laying the groundwork for future applications, we coupled VIC with WRF through the CCSM4 coupler CPL7. In CCSM4, the communication process is separated from the component integration process. All communication processes are performed by Cpl7 and the components run by themselves. Coding work therefore was mainly focused on replacing CLM with VIC. VIC was extracted as it runs in an existing MM5-VIC coupling system for interaction with the flux coupler (because VIC in MM5 is in image mode, i.e., runs at all space for a given time step, as contrasted with point mode, which runs all time steps at a given grid node before proceeding to the next grid node). Based on the VIC version with MM5, we developed our own interface subroutine, which was suitable for the CCSM4 environment to connect coupler and VIC. The WRF-VIC coupling work is essentially completed, with work under other projects on-going to improve the performance.      

Sonessa, M.Y., J.E.  Richey, and D.P. Lettenmaier. (in prep). Evaluation of Water Balance Terms of the SEA with a Land Surface Model and ERA Interim Reanalysis

      The South East Asia (SEA) region includes seven major basins (Mekong, Irrawaddy, Salween, Chao Phraya, Hong (Red ) River, Sittang, Song Ma). Added to the impact of climate and land use changes, there are increasing number of dams being constructed or planned for the near future, especially on the main stream of Mekong and its tributaries. We evaluate the spatial and temporal variability of the water balance terms across SEA using the Variable Infiltration Capacity (VIC) land surface hydrology model, forced by gridded precipitation and temperature, from the ERA-Interim reanalysis and the WRF model.

     The first set of meteorological forcing data for the VIC model, were derived as precipitation (monthly P from the University of Delaware (UDel), adjusted for gauge under catch and for orographic effects, and then to daily from NCEP/NCAR reanalysis), and temperature (monthly CRU, then NCEP/NCAR to daily), The second set of meteorological forcing used as VIC input is based on simulation produced by the Weather Research and Forecasting (WRF) model, applied at 18 km horizontal resolution over Asia to resolve the complex terrain and its influence on meteorological conditions. Soil properties were obtained from the FAO Soil Program. The land use data for the model was obtained from MODIS 2003 (MOD12Q1 Land Cover Product - MODIS/Terra Land Cover 96 Day L3 Global 1 km ISIN Grid - IGBP land cover classification) tiles were acquired through the NASA ECHO website: http://www.echo.nasa.gov/index.html . The ERA interim archive constitutes basically the same land surface variables produced by VIC including surface fluxes of both water and energy, as well as atmospheric moisture flux and storage at multiple levels, which can be vertically integrated to produce a gridded atmospheric water balance. The ERA Interim data cover the period from January 1989 onwards.

 The P from all the three sources, VIC-NCEP, VIC-WRF and ERA-INT, showed similar seasonal pattern as well as comparable magnitudes. P peaked during the northern hemisphere summer season (JJA) contributing about half of mean annual P; while winter (DJF) is the lowest season contributing less than 6 % of mean annual P. As in the case of P, JJA contributes the highest percentage of mean annual ET for all the sources. Even though ET from all the sources showed similar seasonal pattern, the ERA-INT ET is the least variable seasonally with only 11% difference between the peak season (JJA) and the low season (DJF). For the other two sources, VIC-NCEP and VIC-WRF, the difference between the peak and low season in terms of contribution to annual mean ET is 35 %.RF, next to P, is the flux term where all the three sources matched each other both in terms of seasonal pattern as well as flux magnitude. The summer season (JJA) contributes the lion's share of the mean annual RF for all the sources, VIC-NCEP and ERA-INT (52%) and VIC-WRF (44%), while winter season (DJF) contributes the least, VIC-NCEP and ERA-INT (4%) and VIC-WRF (7%). VIC balances the surface water budget by construct; thus, non-closure term for VIC-NCEP and VIC-WRF is less than 1% of annual mean P. However, the non-closure term for ERA-INT is 7.1% of the annual mean P for the SEA region overall. For the seven basins individually, ERA-INT non-closure term ranges from 1.2% for Red basin to 9.8% for Mekong basin. The P and RF have similar spatial pattern with high values in the northwest SEA (Irrawaddy basin) basin and low values near the north and south ends of the SEA. ERA INT P and ET showed less spatial variability relative to those from the other sources. ET increases from north to south annually as well as when divided into four seasons. 

Beyene, T., D.P. Lettenmaier, and J.E Richey. (in prep).  Mekong river reservoir simulation for Current and Future Climate

     Climate change coupled with economic and social development specifically in recent development of construction and operation of dams in the Mekong River basin raises and poses a greater contention and pressure on the limited water resources of the region. In addition to the Chinese cascade of dams, multiple dams are now being proposed for the lower Mekong, as far as Cambodia. The net consequences are subject to extensive debate, but little quantitative analysis. This study investigates the projections of climate change induced hydrologic and water resources implications during the twenty-first century simulated by 20 coupled atmosphere-ocean general circulation models based on the Special Report on Emissions Scenarios A1B and B1. To reduce model bias and uncertainty, ensembles mean (EM) is used for multi-model projections. Although it is difficult to reproduce the present Mekong river discharge in any single model, the EM results produce more accurate representation of the basins river flow. The paper is mainly focused on the state of the art of simulating and optimizing the reservoirs in the upper  and lower Mekong both for retrospective (observed climate) and future climate (IPCC 2007 ) emissions to quantify and asses the combined effect of climate change and reservoir development mainly for hydropower production. The overall approach is to use the VIC macroscale land surface hydrologic model, forced with a regional gridded data set of precipitation, temperature, and wind data. A reservoir simulation and optimization model would be forced using simulated fluxes by VIC. The reservoir model is mainly optimized for maximum hydro power generation.

      Our multi-model median warming by 2070-2099 relative to 1978-2000 was 2.12 and 2.34 o C for A1B and B1 emissions scenario respectively. For the same periods, the models project median precipitation increases of ~ 0.4 and 0.41 m (25 and 27%) with corresponding median runoff changes of  0.25 and 0.27 m (50 and 57%), an increase of ~250,000 mcm, with largest runoff increases resulting mainly from an increase in wet season (May to October) precipitation in all catchments, for A1B and B1 respectively. We perform quantitative analysis of the signatures of the degree of regulation of six upper Mekong reservoirs and 7 lower Mekong reservoirs to downstream riparian countries in term of shifts in seasonal streamflow characteristics. A sharp decrease in wet-season streamflow and an increase in dry-season streamflow were predicted to occur following completion and operation of six dams in the Chinese headwaters and seven lower Mekong dams. Historical degree of regulation of 3% at Gongg and 16% at Jingh in the upper Mekong and  15 % in Pakbeng and 20 % in Latsua are predicted using simultaneously simulating and optimizing  all the dams for aggregated regulation effect in the rivers streamflow.