A34B-01 INVITED
Assimilation of Precipitation and Latent Heating Profiles
Mesoscale reanalysis for the monsoon domain is being recognized as a future need for research on weather and climate. Such a reanalysis, at the present time may require a degree of sophistication that could address the needs of hydrometeorology and the pollution aspects of the monsoon region. This abstract covers one important area for a hydro-meteorological monsoon data base i.e. incorporation of the initial rain in the data assimilation. This FSU data assimilation includes an operational ECMWF (� deg latitude/ longitude) data at 29 vertical levels. To that we add a physical initialization component that includes the following: A reverse cumulus parameterization algorithm that includes satellite and rain gauge based observed rainfall estimates. A reverse surface similarity algorithm that also incorporates the surface rain to provide surface fluxes of moisture consistent with these initial rainfall estimates. An OLR (the outgoing longwave radiation) matching algorithm that provides a close agreement between the model based and the satellite based estimate of OLR. The aforementioned three components of physical initialization reside within a Newtonian nudging that is carried out between hour -36 to hour 0 to provide a rain rate initialization for hour 0. This method carries consistent correlation skills as high as 0.9 for the nowcasting of rain in a regional mesoscale model for each day of analysis. We show that this method does carry a strong positive impact for precipitation forecasts over those of control runs that do not invoke physical initialization. This procedure is also shown to carry a positive impact for mesoscale forecasts out to 3 days into the future. Some comparisons with current one and three dimensional variational data assimilation of precipitation will also be illustrated to show the strength of the Newtonian relaxation based initialization of rain. The reanalysis for the monsoon region may require further degree of sophistication
A34B-02
Asian aerosol assimilation and regional aerosol radiative forcing, using reanalyses, regional models and satellite data
We summarize a state-of-the-art aerosol assimilation effort aimed at recreating the 4D aerosol (including dust and anthropogenic aerosols) heating rates and surface solar fluxes in Asia. This project was launched to better understand aerosols, their radiative forcing and regional climatic impacts in Asia. The PNNL regional model was used to provide meteorology. Satellite and ground based data were integrated for AOD & SSA data. The Iowa aerosol/chemistry model STEM-2K1 assimilated AOD & SSA observations into their model bound by the PNNL meteorology, and produced 2001-2004 aerosol simulations at the $0.45\deg$x$0.4\deg$ resolution. In the vertical there are 23 layers in the troposphere. We used the aerosol simulations and the regional model simulated meteorology to calculate anthropogenic aerosol radiative forcing with an updated Monte-Carlo Aerosol Cloud Radiation (MACR) model. The PNNL regional model is used again with this forcing to assess the climatic effects of Asian anthropogenic aerosols. The 4 year climatological monthly mean anthropogenic aerosol forcing ranges from -20 Wm$^{-2}$ to +10 Wm$^{-2}$ (TOA), from +2 Wm$^{-2}$ to +50 Wm$^{-2}$ (atmosphere) and from -57 Wm$^{-2}$ to -3 Wm$^{-2}$ (surface). In the vertical, the aerosol forcing is mainly concentrated below 650hPa and diminishes slowly towards 200hPa. At the 775hPa level, the forcing often exceeds 1.0 K/day. At this level, the forcing is largest in winter but its interannual variation is larger in fall. Most of the forcing is around South Asia, Southeast Asia, the eastern China and nearby oceans.
A34B-03 INVITED
The Influence of COSMIC Satellite Data on Regional Analysis
The atmospheric limb sounding technique making use of radio signals transmitted by the Global Position System (GPS) has emerged as a promising approach for global atmospheric measurements. As demonstrated by the proof-of-concept GPS Meteorology (GPS/MET) experiment and more recently by the CHAMP and SAC-C missions, the GPS radio occultation (RO) sounding data are of high accuracy and high vertical resolution. On 15 April 2006, the joint U.S.-Taiwan COSMIC/FORMOSAT-3 mission, a constellation of six microsatellites, was launched from the Vandenberg Air Force Base. These satellites are being deployed to their final orbits, which would take about a year. During the early phase of the deployment, the satellites are closely located. This offers a unique opportunity to examine the precision of the GPS RO measurements. The COSMIC data are available in near real-time for global weather analysis and prediction and for climate monitoring. Currently, COSMIC is producing approximately 1300 GPS RO soundings per day at the end of August 2006. This number will be increased as the satellites are further separated through the deployment process. Radio occultation measures phase and amplitude of the microwave signals emitted from GPS. These signals are inverted to obtain profiles of signal bending, atmospheric refractivity, pressure temperature and water vapor. The main objective of the COSMIC/FORMOSAT-3 mission is to demonstrate the value of these radio occultation products for weather forecasting, climate monitoring, ionospheric research and space weather prediction. This presentation will provide an overview of the COSMIC/FORMOSAT-3 program. We will present results on the influence of COSMIC data on the regional analysis over the data void regions, particularly over the tropics and high latitudes. For further information on the COSMIC/FORMOSAT-3, please refer to http://www.cosmic.ucar.edu/.
http://www.cosmic.ucar.edu/
A34B-04 INVITED
Monsoon variability and hydrology of South Asia
With the prospect of a changing climate, a major concern is the future availability of water for agriculture, irrigation, power generation and human consumption. Nowhere is the concern greater than in South Asia where, in addition to an uncertain future climate, is the prospect of a rapidly expanding population. It is possible to identify two situations: regions that are not irrigated and regions that are irrigated using water from major rivers. For the former region, predictions of future precipitation patterns will suffice. But for the irrigated regions, predicting regional rainfall is not sufficient because water availability depends on precipitation from all parts of the catchment. In addition there is a need to marry rainfall variability (remote and local) with the hydrology of the major rivers themselves. Finally, there is the imperative question of what the future rainfall patterns will be like and what aspect of climate may determine them. First we consider the present climate and relationship of the river discharge of major systems to various temporal scales of motion. In particular, we note the importance of intraseasonal variability in promoting both regional rainfall patterns and in producing a proportion of interannual variability. We also consider forced interannual variability through variations in sea-surface temperature patterns. For rivers such as the Ganges, the variability associated with ENSO is very strong. For the Brahmaputra, SST relationships are much weaker although there is some evidence that winter mid-latitude anomalies may be associated with Himalayan-Tibetan snow pack and therefore early stream flow in late spring. Second, we consider the results of the IPCC coupled ocean atmosphere model integrations. We use quantile-to-quantile rainfall corrections between model and observations to determine present climate river discharges. These simulations are then applied to future discharges using the IPCC models simulations for the next 100 years. The purpose of the study is to determine the ability of models to simulate present river discharge and, thus, their ability to determine future states. In addition, the exercise is useful in determining what hydrological quantities are useful in reanalysis efforts.
A34B-05 INVITED
Regional Modeling Over Complex Terrains: Hydroclimate Challenges Over South Asia
The climate of South Asia is greatly influenced by the monsoon circulation and topographic forcing represented by complex terrain features including the massive Tibetan Plateau, and narrow mountains such as the western Ghats, Arakan Yoma, and Daiwna-Bilauktaung. Observed rainfall and outgoing longwave radiation show that mountains anchor convection, which lead to intense precipitation on the windward side or further upstream. The impacts of the diurnally varying atmospheric heating associated with these convection centers, particulary those related to narrow mountains, are not well understood. To advance our understanding of the regional energy and water cycle in South Asia, the effects of the Tibetan Plateau and mesoscale mountains must be assessed and realistically simulated. Recently, the Penn State/NCAR Mesoscale Model (MM5) has been used to develop a regional analysis of the hydroclimate of South Asia. The model was applied at 50 km horizontal resolution, with data assimilation to constrain the large scale circulation. Temperature and winds from the NCEP/NCAR global reanalysis were spatially interpolated to the MM5 grids and assimilated in the regional simulation using simple nudging. The simulation was performed for 1998 � 2004. Evaluation and analysis of the simulation will be presented, focusing on the regional hydrological cycle and effects of the complex terrain, to provide guidance to the retrospective analysis being planned for South Asia.
A34B-06 INVITED
Influence of Aerosols on Monsoon Circulation and Hydroclimate
Long recognized as a major environmental hazard, aerosol is now known to have strong impacts on both regional and global water cycles and climate change. In the Asian monsoon regions, the response of the regional water cycle and climate to aerosol forcing is very complex, not only because of presence of diverse mix of aerosol species with vastly different radiative properties, but also because the monsoon is strongly influenced by ocean and land surface processes, land use, land change, as well as regional and global greenhouse warming effects. Thus, sorting out the impacts of aerosol forcing, and interaction with the monsoon water cycle is a very challenging problem. Up to now, besides the general notion that aerosols may significantly impact monsoon through altering large scale radiative heating gradients, there has been very little information regarding the specific signatures, and mechanisms of aerosol-monsoon water cycle interaction. In this talk, based on preliminary results from observations and climate model experiments, I will offer some insights into how aerosols may impact the Asian monsoon water cycle, in particular the effects of absorbing aerosols (dust and black carbon), and the role of the Tibetan Plateau. The influence of aerosol forcing relative to those due to sea surface temperature and land surface processes, and impact on potential predictability of the monsoon climate system will also be discussed.
A34B-07
Lessons From the NCEP North American Regional Reanalysis Project
The NCEP North American Regional Reanalysis (NARR) project had a clear foremost objective: to create a long- term, consistent, high resolution climate dataset for the North American domain as a major improvement upon the earlier global reanalyses in both resolution and accuracy. The assessment of the authors of the NARR AMS Bulletin paper was that this objective has been fully met. Precipitation assimilation, the first in a reanalysis project, resulted in precipitation fields very near those of the ingested precipitation analyses, ensuring that over regions with reasonable density of gauge observations, the hydrological cycle is more realistic than if the model was free to forecast precipitation. With respect to fits to data, not only have the near-surface temperatures and winds been shown to be closer to the observations than those of the NCEP/DOE Global Reanalysis (GR2), as one would expect, but very substantial improvements in the accuracy of winds and temperatures throughout the troposphere compared to that of GR2 have been demonstrated as well. In regard to lessons that may benefit future regional reanalysis projects, three questions deserve attention. First, what are the features of the NARR system that are responsible for its success, to the extent their contributions can be assessed. Next, what are the disappointments, including features that were expected to be beneficial but failed to be confirmed as such, and why. And finally, what are the weaknesses that would or should have been addressed had there been more manpower and time to do so. The first two questions have been addressed in the published BAMS paper and will be briefly summarized in the presentation. As to the third question, a number of weaknesses come to mind. For example, the Eta 3DVAR data assimilation system (EDAS), as used in NARR, 1) achieved little or no benefit from assimilating near- surface data, 2) was resulting in considerable gravity-wave activity, and 3) derived no discernible benefit from its direct assimilation of radiances. There are clear options for addressing each of these issues. Additionally, while precipitation assimilation performed extremely well over areas having a reasonable spatial density of precipitation gauge observations, such as CONUS, other areas of sparse availability of gauge observations might have done better with the model predicted precipitation instead. Also, while tests of the assimilation of synthetic hurricane winds looked quite promising, there was insufficient time for adequate testing prior to production and hence such assimilation was not used. A number of promising and well-understood candidate model refinements were not used for lack of suitable time for testing. Every project of course has to face constraints of resources and time, but a judicious use of past experience in addressing the choices at hand should go far in achieving the optimal result given the state of technology available to the project.