报告题目：Development and Applications of North American Land Data Assimilation System at NCEP: Past, Present, and Future
报 告 人：Dr. Youlong Xia
单 位：NOAA/NCEP/EMC
时 间：2016年8月15日上午10:00-11:00
地 点：大气所40号楼319会议室
We present the history, current status, and future plans of the purpose, methodology, and product suite of NLDAS. The purpose of NLDAS is to support numerous research and operational applications in the land modeling and water resources management communities, including drought monitoring. The NLDAS (http://www.emc.ncep.noaa.gov/mmb/nldas/; http://ldas.gsfc.nasa.gov/nldas) is a multi-institutional collaborative project sponsored by NOAA’s Climate Program Office and NASA’s Terrestrial Hydrological Program. Since the NLDAS was initiated in 2000, it has been conducted in two phases. Phase 1 initiatives spanned from 2000 to 2005 (NLDAS-1) and its objectives were to establish the NLDAS configuration (including collection of soil, vegetation, albedo, and terrain data, etc.), initial selection of land-surface models (LSMs), generation of surface forcing data sets, and production of continuously cycled retrospective model runs for a 3-year period (from October 1996 to September 1999), with evaluation/validation of model output. Phase 2 is an extension of Phase 1 and its initiatives include: A) much longer retrospective simulations from January 1979 through 2009 executed during the 2006 to 2009 time frame, followed by: B) the first NLDAS-2 real-time quasi-operational simulations, including extensive product evaluation and pilot-product application from 2010 to 2013. In August 2014, NLDAS-2 was formally implemented into official NCEP operations (http://www.nco.ncep.noaa.gov/pmb/products/nldas/).
NLDAS has a long successful history of producing soil moisture, snow cover, runoff and streamflow products via application of surface meteorology and precipitation datasets to drive multiple land-surface models (LSMs). These model outputs have been comprehensively evaluated against in situ observations, satellite retrievals, and reanalysis products. NLDAS products are being used by over 5,000 users (http://ldas.gsfc.nasa.gov/nldas/NLDASnews.php) worldwide, including academia, governmental agencies, and private enterprises. Over 44 million files and 93 Tb of data were downloaded in 2014 from NASA alone. In particular, NLDAS products are supporting U.S. operational drought monitoring and prediction tasks including the U.S. Drought Monitor, CPC seasonal Drought Outlook, and other National Integrated Drought Information System activities (NIDIS, drought.gov). As we endeavor to increase the quality and breadth NLDAS products, more and more benefits will be gained from this system.
We recognize that various improvements need to be made in the current operational NLDAS-2 system. In particular, in operational practice we are thus far missing the actual data assimilation of either: 1) realtime satellite-based land surface states (e.g. soil moisture, snow cover, vegetation cover, LST, terrestrial water storage, other), or: 2) realtime in situ data such as streamflow observations. Additionally, the land modeling research community has made substantial progress in improving land model performance and skill through adding/improving many realistic physical processes, obtaining more realistic model parameters, and enhancing the quality of surface meteorological forcing, in particular, science advance and understanding in model physical processes such as surface exchange coefficient, dynamic vegetation, irrigation/pumping, groundwater dynamics, and total runoff partitioning, and new vegetation/soil datasets such as realtime vegetation cover (including burn areas), vegetation type and soil classes. These upgrades need to be incorporated in the current NLDAS-2 system. We also need to improve the spatial resolution of NLDAS-2 from its current medium resolution (0.125-degree) to extra high spatial resolution (0.03125-degree) as well as expanding the domain to include all of North America. In line with current NCEP thrusts toward a unified modeling strategy, the NLDAS and its global counterpart GLDAS (NCEP Global Land Data Assimilation System) will be merged into a single unified NCEP Land Data Assimilation System. This new LDAS system will be run within the NASA Land Information System (LIS) software framework, including actual data assimilation processes with different domains and spatial resolutions. Links to the new NWS National Water Center (NWC) National Water Model (NWM) will be leveraged to more comprehensively support regional and global drought monitoring and forecasting requirements, as well as provide the initial land states for NCEP regional models, anticipating an integrated land-hydrology approach for global coupled weather and climate forecast systems.
Finally, the NLDAS prediction mode – Seasonal Hydrological Forecast System jointly developed by Princeton University and University of Washington at Seattle, will briefly discussed. Although it was run at NCEP only as an experimental and quasi-operational forecast system, as this system is comprehensively evaluated, improved and expanded, it will be operationally implemented in future.
