Climate

Climate Image

How much is it raining, where?

   “Climate” is a primary driver of what happens in a river basin. How hard is it running? How hot is it? How much rain leads to flooding? Is unusually “low water)” in front of Phnom Penh a result of withholding water behind dams in the upper Mekong (as some newspapers will attest), or is it a result of the fact that it simply didn’t rain in the upper basin?

   The general aspects of the climate of the Mekong basin are described elsewhere on this site. Of immediate interest is how to rigorously describe this climate, and then how to incorporate it into models that take climate data, and transform that information into hydrology models. Such models typically require daily precipitation, maximum and minimum temperature, and mean wind speed. Other forcing data (e.g. downward solar and longwave radiation, humidity) are derived from the daily temperature or temperature range. Here three sets of such climate forcings are described, as developed in Sonessa et al (in prep)

   The first set of meteorological forcing data for the VIC model, precipitation, maximum temperature, minimum temperature and wind speed are prepared as follows. Monthly precipitation (P) was obtained from the University of Delaware (UDel) Willmott and Matsura (2006) gridded observation. Gauge Precipitation adjusted for gauge under catch as described by Adam and Lettenmaier (2003) and for orographic effects as described by Adam et al. (2006). Monthly gridded minimum and maximum temperatures were obtained the Climate Research Unit (CRU) of the University of Eastern Anglia (Mitchell et al., 2004). Wind speed data was obtained from NCEP-NCAR reanalysis (Kalnay et al. 1996).

Annual Precipitation for Modeling   

   The daily variability of NCEP/NCAR is used to create daily P and temperatures data using monthly CRU (for temperature) and Udel (for P) data as a control. Thus, for a given month the daily precipitation varies like the NCEP/NCAR data while the amount is controlled by (add up to) that month's UDel precipitation. The data prepared using the above mentioned steps are of 0.5 degree resolution. These data were interpolated to 1/12 grid cells using inverse distance interpolation. Besides, temperature data were interpolated using a lapse rate of -6.5 °C per km to adjust temperature from the 0.5 degree grid cell to each elevation of the 1/12 grid cell. The VIC outputs using these forcing data is termed as VIC-NCEP.

   The second set of meteorological forcing used as VIC input is based on simulation produced by the Weather Research and Forecasting (WRF) model. WRF is a community mesoscale atmospheric model designed for use on regional grids (Skamarock et al. 2008). It includes a fully compressible non-hydrostatic atmospheric dynamical core with one-way and two-way nesting capability and a full suite of physics. WRF was applied at 18 km horizontal resolution over Asia to resolve the complex terrain and its influence on meteorological conditions. The simulation was initialized on January 1, 1997 and last through December 31, 2008. The WRF simulation was driven by global circulation (winds, temperature, and moisture) from the NCEP/NCAR global reanalysis (Kalnay et al. 1996) at 6 hours intervals by assimilating the large-scale conditions using a relaxation treatment of the lateral boundary conditions. In addition, spectral nudging was applied to constrain the WRF simulation to follow closely the NCEP/NCAR large-scale circulation throughout the whole WRF model domain. Because of the stronger large-scale control, regional simulations generated using interior nudging can simulate more realistic spatial and temporal variability compared to regional simulations driven only by large-scale conditions at the lateral boundaries (e.g., von Storch et al. 2000; Leung et al. 2004; Miguez-Macho et al. 2004).

   The WRF simulation was performed using the WSM6 cloud microphysics scheme, Grell convective parameterization, the RRTMG scheme for shortwave and longwave radiation, and the Mellor-Yamada-Janjic boundary layer scheme for atmospheric processes (Skamarock et al. 2008). The Noah land surface model (Chen and Dudhia 2001) was used to simulate land surface processes that interact with the atmosphere. Model outputs were archived every three hours. After one year of model spinup, daily maximum and minimum surface temperature, precipitation, and 10-m winds were provided as atmospheric forcing for VIC to perform simulations from 1998 – 2006.

   The ERA-Interim (Dee et al. 2011) project was initiated by the European Center for Medium Range Weather Forecasting (ECMWF). It is a reanalysis of the global atmosphere covering the data-rich period since 1989, and continuing in real time. The ERA-Interim project was started to provide a bridge between ECMWF’s ERA-40 (Upsala et al. 2005) reanalysis which runs 1957 through 2002 and next generation extended reanalysis envisaged at ECMWF. The ERA Interim attempts to improve certain key aspects of ERA-40. These aspects include: the hydrological cycle representation, the quality of the stratospheric circulation, and the handling of biases and changes in the observing system. The main developments in the ERA-Interim data assimilation compared to ERA-40 to result in the above improvements are: 12 hour 4D-Variatioanl analysis, T255 horizontal resolution, better formulation of background error constraint, a revised humidity analysis, improved model physics, data quality control that draws on experience from ERA-40 and JRA-25, variational bias correction of satellite radiance data, and other improvements in bias handling, more extensive use of radiances, and improved fast radiative transfer model.

   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. We used 9 years of monthly averages of precipitation, evapotranspiration, runoff the period 1998 to 2006. These variables are derived entirely from the ECMWF data assimilation system and have no direct relationship to observations. The archived ERA Interim gridded data were interpolated to 0.25 º grid using an inverse distance interpolation. Atmospheric and terrestrial water balance terms expressions are well explained by Su and Lettenmaier (2009) and Su et al (2006). The water balance terms obtained from the ERA-Interim are referred to as ERA-INT in this paper. For instance the ERA-Interim precipitation is referred to as ERA-INT P.