Skip Top Navigation
only rendered in printing; banner identifies JCSDA logo and the NOAA, NASA, USAF, and US Navy organizations
JCSDA Logo and banner

JCSDA News and Newsletters

Center News

Science Update - Air Force Weather Agency Land Information System and Snow Analysis Model

The Air Force Weather Agency (AFWA) and NASA Goddard Hydrological Sciences Branch have worked together extensively over the past few years to build a system to support high resolution Department of Defense (DoD) land surface characterization capabilities through the NASA Land Information System (LIS). The most significant recent advancement in LIS is the inclusion of a generalized version of the NASA Global Modeling and Assimilation Office (GMAO) Ensemble Kalman Filter (EnKF) to support future data assimilation goals, including remotely sensed surface observations. The GMAO EnKF module was released in LIS version 5, and is available for download from the NASA LIS website at http://lis.gsfc.nasa.gov.

At left: MODIS/AMSR-E Snow Cover; at right: AMSR-E Snow Depth; (both 2-20-2008)

At left: MODIS/AMSR-E Snow Cover
At right: AMSR-E Snow Depth
(both 2-20-2008)

The AFWA continues to participate with the National Centers for Environmental Prediction (NCEP) Environmental Modeling Center partnership to improve LIS capabilities. Currently, the LIS team is working with the JCSDA and NCEP to build an interface to the Community Radiative Transfer Model. The combination of the EnKF and CRTM modules in LIS will enable a much broader satellite-based land surface observation assimilation capability in the future. Additionally, NASA is nearing the completion of a project to evaluate the assimilation of geostationary-based skin temperature observations and NASA Earth Observing System (EOS) Moderate resolution Imaging Spectroradiometer (MODIS) snow cover and Advanced Microwave Scanning Radiometer for EOS (AMSR-E) snow water equivalent observations into LIS through the EnKF module.

AFWA is actively working to complete final software engineering work prior to operational implementation of LIS by late 2008. The LIS software will initially be implemented in parallel with the current AFWA Agriculture Meteorology (AGRMET) model for a short period of time until downstream users and applications complete the transition to use LIS-based products.

AFWA is also working with the NASA Goddard Snow Team (GOST) to add new capabilities to incorporate remotely sensed snow cover and depth observations into the AFWA global Snow Depth Analysis Model (SNODEP). The current operational SNODEP model relies primarily upon surface-based synoptic snow depth measurements to create the global snow depth analyses. The improved science developed by the GOST team is a new algorithm that blends MODIS snow cover measurements with AMSR-E snow depth and snow cover measurements. The new algorithm, termed ANSA for AFWA NASA Snow Algorithm, provides a significant boost in the AFWA global snow measuring mission (see figure above).

AFWA is actively working to integrate the new ANSA capabilities into the SNODEP model with a target operational date in the summer of 2009. NASA and AFWA continue to collaborate on further development of the ANSA algorithms to add new snow melt products using AMSR-E and NASA QuikSCAT data. The new products should help to augment the AFWA global snow mapping and snow characterization mission support to the DoD community.

J. Eylander, AFWA; C. D. Peters-Lidard, D. Hall, J. Foster, and R. Reichle, NASA GSFC; and S. Kumar, SAIC Inc. and NASA GSFC



Science Update - The Community Radiative Transfer Model

A major achievement of the JCSDA has been the development of a Community Radiative Transfer Model (CRTM), portions of which are now used by all the partner agencies. The CRTM makes possible the rapid assimilation of millions of satellite observations at over a thousand spectral wavelengths from dozens of different instruments each day.

CRTM diagram

A rapid, accurate radiative transfer model is one of the keys to success in assimilating satellite observed radiances. Realizing its importance, the JCSDA, at its inception in 2002, initiated the construction of a Community Radiative Transfer Model (CRTM), engaging internal and external developers.

The CRTM is a software library for computing satellite instrument radiances from atmospheric and surface state variables. It emphasizes modularity and code reuse and is independent of computing platform or assimilation system. Its current capabilities include Forward, Tangent-linear, Adjoint and Jacobian models for assimilation of clear and cloudy IR and MW radiance observations from satellites over ocean, land, snow, and ice surfaces.

CRTM modules include: Gaseous absorption, including H2O, O3 (model variables), CO2, CH4, N2O, CO, and O2 (gas concentrations specified); Atmospheric aerosols for eight different aerosol types; Clouds and precipitation for liquid phase (liquid water and rain) and solid phase (ice, snow, graupel, and hail); Microwave land and ocean emissivity; and Infrared land and ocean emissivity.

While current assimilations are limited to observations from clear regions of the atmosphere, the CRTM is ready for integration into assimilation schemes for cloudy and precipitating areas as these are developed. It is also ready to be coupled with air quality and NWP models that treat CO2, CH4, N2O, and CO as variable gasses. This capability will enable, for example, monitoring of greenhouse gases within the context of an NWP model based on observations from hyperspectral IR sounders such as AIRS and IASI.

Fuzhong Weng, JCSDA



Impact of Satellite Altimetry on JCSDA Ocean Data Assimilation and Seasonal Climate Forecasts

(From the March 2008 JCSDA Quarterly)

Since the ocean provides a significant memory for the climate system, a critical element in climate forecasting with coupled models is the initialization of the ocean with states from an ocean data assimilation system (ODAS). Since October 1992 global ocean surface topography has been observed with TOPEX/Poseidon (1992-2005) and Jason-1 (2001-present) altimeters, both joint NASA/CNES missions. These satellites monitor changes in ocean heat storage and ocean currents.

Figure 1.

Figure 1. Impact of altimetry assimilation on GODAS state estimates is assessed through the RMS differences (cm) from Topex/Poseidon and Jason-1 SSH anomalies for 1993 to 2007. The right- (left-) hand figure shows the RMS difference with (without) altimetry assimilation.


Figure 2.

Figure 2. The anomaly correlation skill score for heat content in the upper 300 m from the GMAO CGCMv1 for 6-month forecasts from 1 July initial conditions. The ocean is initialized from the EnKF with (left panel) and without (right panel) assimilation of satellite SSH anomalies. Only correlations higher than 0.6 are shown. The forecasts are validated against their own analyses.


Both NOAA/NCEP and the NASA/GMAO use sea surface height (SSH) anomalies from these altimeters in their ODAS with the goal of improving global ocean state estimates and also seasonal climate forecast skill. The NCEP global ocean data assimilation system (GODAS), which currently provides initial conditions for the NCEP coupled Climate Forecast System (CFS), uses 3dVAR with the GFDL MOMv3. The GMAO system uses an Ensemble Kalman Filter (EnKF) with the Poseidon ocean model to initialize their CGCMv1. Both systems, although global, focus on the tropical oceans. In addition to the altimetry data, which provides information only at the surface, the ODAS assimilates temperature profiles from XBTs, fixed tropical moorings (TAO, TRITON, and PIRATA arrays) and the global Argo array.

Both assimilation methods are designed to modify the mass field of the ocean model through corrections to temperature and salinity. Differences between the model SSH and observed SSH are translated into corrections to the temperature and salinity throughout the water column through the specification of background error covariances.

Figure 1 shows the RMS differences between the altimeter observations and the GODAS dynamic heights. The area of low RMS differences (grey regions) is increased substantially with the assimilation of the altimeter data. In the tropics the RMS differences remain somewhat larger (4-5 cm) in the region of the tropical instability waves and the recirculation of the Brazil current. Outside of the tropics in the Gulf Stream and Kuroshio, which are not well resolved by climate-scale models like GODAS, the RMS differences are larger still.

The GMAO's ODAS, the EnKF with online bias correction, has also been used to initialize seasonal forecasts with and without assimilation of altimeter data. As for other coupled models, the forecast skill varies seasonally. It is difficult to discern significant differences in skill from the different ocean initializations for January starts. The skill for July starts is longer-lived and there are discernable differences in performance for the two ocean initializations. Figure 2 shows that the skill of 6-month forecasts of upper-ocean heat content in the tropical oceans is improved with the assimilation of SSH anomalies.

We are now anticipating the joint NOAA/NASA/CNES/EUMETSAT Ocean Surface Topography Mission (OSTM), or Jason-2, which will be launched in June 2008, to extend the time series of sea surface topography measurements to two decades.

(David Behringer, NOAA/NCEP/Environmental Modeling Center, and Michele Rienecker, NASA/GSFC/GMAO)

JCSDA Quarterly Newsletter

Please send news items for the JCSDA Quarterly to:
george.ohring@noaa.gov

Deadline:
2 weeks before the end of each calendar quarter.
Last modified on October 8, 2008 12:10 PM
NOAA | NWS | NASA
Navy Weather | Air Force Weather
Privacy |  Link & Product Disclaimers | Information Quality
Accessibility | Contact JCSDA webmaster
Hosted for JCSDA at STAR
Level A conformance icon, W3C-WAI Web Content Accessibility Guidelines 1.0Valid HTML 4.01