Loading...
Please wait, while we are loading the content...
Similar Documents
Megha-Tropiques satellite and the Advanced Microwave Scanning Radiometer-2 (AMSR-2) instrument on the Japanese Global Change Observation Mission – Water (GCOM-W1) satellite, as well as operational sensors, such as the Special Sensor Microwave Imager/Sounder (SSMIS), Advance Microwave Sounding Unit (
| Content Provider | Semantic Scholar |
|---|---|
| Copyright Year | 2015 |
| Abstract | In our previous proposal we introduced a method to address the problem that has plagued the assimilation of precipitation: the model rains as observed until the assimilation ends, and then it “forgets” the precipitation assimilation and after a few hours, returns to the original forecast. Our proposed approach is based on 1) Use of an Ensemble Kalman Filter (the LETKF) which modifies all the model variables (including potential vorticity), not just the moisture and temperature in the vertical column, as done in current systems. This update of dynamic variables should make the model “remember” the assimilation of precipitation during the forecast. 2) The LETKF, like all data assimilation methods, assumes that the observation errors are Gaussian, which is false for precipitation. To address this we proposed to use a Gaussian Transform on both the rain observations and the model precipitation. We showed in Lien et al (2013, Tellus) with a perfect model OSSE that the new approach of assimilating precipitation was very successful, achieving for the first time a significant improvement in the 5-day forecast. We then assimilated TMPA real precipitation with a low resolution NCEP/GFS model, and found that the model and the observations have very different statistics, but that assimilating precipitation also improved the 5-day forecast for all variables and all regions of the world (Lien et al., 2015a, 2015b, MWR, under revision). After these successful results we applied Ensemble Forecast Sensitivity to Observations (EFSO, Kalnay et al, 2012, Hotta et al. 2015, to be submitted in July 2015 to MWR) to determine whether each TMPA observation improves the analysis or makes it worse. The first experiment showed that 1) The Gaussian Transform works much better than all the alternative methods of No Transform, No Assimilation of rain, and Assimilation with the widely used Logarithmic Transform; b) The TMPA observations were more useful reducing forecast errors over the ocean and less useful over land, especially over regions of low precipitation. We propose to extend our initial successful results in several ways so that at the end of three years the system for effective assimilation of IMERG precipitation will be ready for operational testing for implementation. The tests we will perform include: a) Use IMERG and compare with TMPA. b) Use a higher resolution GFS model with assimilation of current observing systems. c) Test using a Gaussian Transform for the model precipitation based purely on the ensemble. d) Assimilate precipitation in hurricanes, using GPM radar observations, to test whether the track and intensity forecasts are improved. e) Implement EFSO and Proactive QC to assimilate the optimal set of IMERG observations. f) Compare the impact of assimilating early IMERG with final IMERG estimates. g) Understand why the TMPA observations are found (using EFSO) to be more useful over ocean than over land, and whether this is related to convective precipitation. h) Explore whether using lightning information can improve the assimilation of convective precipitation. i) Use EFSO to estimate IMERG/TMPA observation errors. We will also collaborate with the team at RIKEN, Japan, led by Prof. Takemasa Miyoshi, which will perform similar data assimilation of the Global Satellite Mapping of Precipitation (GSMaP), a product analogous to TMPA. This team will use the Nonhydrostatic Icosahedral Atmospheric Model (NICAM). This should provide useful comparisons of TMPA and GSMaP and the methodologies used. Christopher Kidd/University of Maryland, College Park A Physically-Based Scheme for the Retrieval of Precipitation from Cross-Track Sensors in the GPM Constellation The retrieval of precipitation from all available Earth observations sensors is critical to achieve the spatial and temporal sampling required to capture and represent the precipitation across the Earth’s surface. The Global Precipitation Measurement mission Core Observatory provides a keystone in the cross-calibration of observations and precipitation retrievals for a host of constellation satellites. The current GPM constellation includes not only 6 conically-scanning (CS) passive microwave (PM) radiometers, but also 6 cross-track (XT) PM radiometers, therefore the inclusion of precipitation estimates from the latter play a crucial role in providing comprehensive observations. An initial version of a physically-based retrieval scheme for the XT sensors has been developed and is currently implemented for the operational production of precipitation products at the Precipitation Processing System at NASA’s Goddard Space Flight Center. The scheme is built upon the Goddard Profiling scheme that is used for retrievals from CS observations, but with modifications to account for the XT characteristics of the instrumentation. At the heart of the GPROF scheme is the database that is used as a reference to select comparable atmospheric profiles to those being observed by the sensors; the CS database is built upon a set of actual observations matched against surface/satellite reference data together with profiles from NASA’s Multi-scale Modeling Framework (MMF). Initial testing of the XT scheme implementing a similar database structure revealed problems with the representativeness of the database that resulted in significant regions of ‘missing’ retrievals. Consequently, the current version of the database for the XT retrievals rely solely the MMF model to generate the database entries, although the final database is bias-corrected against the CS database to ensure consistency across the different instruments. Initial comparisons of retrievals from both the CS and XT sensors has been carried out over the US and Western Europe. These results show that the XT results are comparable with the results from the CS results. Indeed, in many cases the XT retrievals perform better than the CS retrievals when considering correlation and root mean squared error statistics. However, the XT performance is slightly less well in terms of skill score due to poorer detection, but better false-alarm occurrences. Therefore, this project aims to build upon the success of the XT MMF-based database scheme to further refine and improve precipitation estimates from these sensors. In particular, the proposed project will include addressing the following issues: i) testing and evaluation of the current MMF-based retrieval scheme to identify situations where retrievals differ from the CS scheme and other precipitation products; causes of the discrepancies will be investigated and feed back to the MMF development; ii) assessing and quantifying dependences of the retrieval scheme upon sensor-dependent characteristics. Scan position dependency is currently accounted for through postretrieval correction; changes in earth incidence angle affect resolution, polarization and atmospheric path; iii) representativeness of the single (all surfaces, all scan positions) database generated by the MMF model. Modifications include better MMF-observational comparisons, particularly with the GPM core satellite (GMI, DPR and combined) products; and, iv) processing efficiency of the retrieval scheme, from the database generation, preprocessing system and the actual retrieval itself. In particular, can the database be adequately represented by fewer profiles without degradation in performance? This research will contribute greatly to the overall aims and objectives of NASA’s Precipitation Measurement Mission by providing precipitation estimates from operational instrumentation and to the wider scientific community. Min-Jeong Kim/Morgan State University All-Sky GPM Microwave Imager (GMI) Radiance Data Assimilation Global Products from the GEOS-5 System in Support of the GPM Mission We propose contributions to the PMM science team by developing (1) global modeling and assimilation methodologies to utilize all-sky GMI radiance data in the Goddard Earth Observing System (GEOS-5) 4D-EnsVar atmospheric data assimilation system (ADAS), (2) prototypes of all-sky GMI radiance data assimilated global atmospheric and surface analysis products and (3) global downscaling methodologies to produce fine scale precipitation products based on GEOS-5 analyses that actively assimilate GMI radiance data, along with millions of other satellite and conventional observations currently utilized to produce near real time global weather forecasts at the NASA Global Modeling and Assimilation Office (GMAO). This effort builds upon an ongoing, funded research project at the GMAO to assimilate cloudand precipitation-affected satellite radiance data in GEOS-5 system. These proposed developments will extend the existing assimilation framework for all-sky GMI radiance data over the ocean to all-sky GMI data over land by enhancing methodologies to consider surface contributions properly in the observation operator during the analysis process so that information from GMI observations can be projected in the analysis to improve precipitation forecasts over land. It is expected that assimilating the GMI all-sky radiance data will improve the GEOS-5 atmospheric analyses and will complement GEOS-5 land surface analyses. The GMI instrument’s wide range of frequencies will improve GEOS-5 precipitation analysis and forecasts, especially over land. The improved analysis of precipitation will enhance the inputs for the GEOS-5 land surface model, which simulates hydrological processes. The quality of land-surface model fields such as skin temperature, soil moisture and snow coverage are critical for maximizing the impacts of microwave radiance data on the atmospheric analyses. By leveraging GMI radiance data assimilation components that directly impact the atmosphere and land, we propose to investigate the development of GEOS-5-derived analysis products and downscaled precipitation products. The aforementioned efforts will support the original goal of the GPM mission to e |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://nspires.nasaprs.com/external/viewrepositorydocument/cmdocumentid=491706/solicitationId=%7B94547DCE-2787-06A5-BF5D-20F5EDDAB3DF%7D/viewSolicitationDocument=1/PMM15%20selections.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
