Past 2012 JCSDA Seminars

The NPP Suomi satellite has 3 sensors, ATMS, CrIS, and OMPS which are of high interest for assimilation into the Navy's 4-Dimensional Variational (4D-Var) assimilation system, NAVDAS-AR. This presentation will primarily address the work done with the ATMS sensor, and will show some of the early work with OMPS; however, the assimilation of CrIS will be saved for a future date. The core of the Navy's global forecast system is the 42-level NOGAPS model, the follow-on system is the 50-level NAVGEM model both with a model top of 0.04 hPa and both are accompanied by the 4D-Var system NAVDAS-AR. The system currently assimilates radiances from microwave (AMSU-A, MHS, SSMIS) and infrared satellite sounders (AIRS and IASI) in addition to the bending angle from GNSS-RO. The current NAVDAS-AR system can compute bias corrections for satellite radiances from either an offline Harris-Kelley type approach, or using a variational bias correction (varBC). The experiments using the 42-level NOGAPS model use a Harris-Kelly bias correction, while NAVGEM experiments use the varBC. The assimilation of ATMS produces stable innovations (observation minus background) over time, with biases of similar magnitude to those of heritage MW sounders. The current observational system has good data coverage in space and time from IR/MW sounders; thus a forecast metric such as 500 hPa anomaly correlation against self-analysis does not show significant impact from the addition of ATMS assimilation. However, the examination of adjoint-based observation impact using a 24-hour global moist total energy error norm shows consistent positive impact from ATMS with very similar observation impact as seen in AMSU-A.

Spaceborne passive microwave observations from sensors such as AMSU and ATMS provide temperature and moisture sounding in both clear air and non-precipitating clouds, along with indirect determination of rain rate over heavier precipitating clouds. Improvements in mesoscale forecasting could be anticipated if such observations were made with higher spatial and temporal resolution, as possible, for example, from either geostationary orbit or from a fleet of LEO satellites. While both of these mission concepts are being studied, there is an associated need to understand how data from such new observation systems might be used within radiance assimilation schemes. We discuss in this talk progress in extended Kalman filtering and radiative transfer modeling applied to the problem of how to optimally use high time-space resolution passive microwave imagery, and specifically, the potential for ?precipitation locking? of NWP models onto passive microwave data. It is suggested from ab initio precipitation locking studies, although not yet proven, that time and space scales of ~15 minutes and ~15 km are needed to maintain small enough model errors in all fields so as to keep NWP models locked under conditions of mesoscale convection. The concept of locking will be illustrated through geostationary microwave observation system simulation experiments. New developments in fast Jacobian RT modeling and CubeSat based sensor concepts for facilitating high temporal/spatial measurements will also be discussed.

A conundrum of predictability research is that while the prediction of flow dependent error distributions is one of its main foci, chaos fundamentally hides flow dependent forecast error distributions from empirical observation. Empirical estimation of such error distributions requires that one obtain a large sample of error realizations given the same flow and the same observational network. However, chaotic elements of the flow and the observing network make it practically impossible to observe and collect the conditioned sample of errors required to empirically define such distributions and their variance. These variances are ?hidden?. Here, an exposition of the problem is developed from an ensemble Kalman filter data assimilation system applied to a 10 variable non-linear chaotic model and 25,000 replicate models. The output from this system motivates a new analytical model for the distribution of true error variances given an imperfect ensemble variance. This model is defined by 6 parameters that also determine the optimal weights for the static and flow dependent parts of Hybrid error variance models. Six new equations enable these hidden parameters to be accurately estimated from a long time series of (innovation, ensemble variance) data pairs. This new-found ability to estimate hidden parameters provides new tools for assessing the quality of ensemble forecasts, tuning Hybrid error variance models and for post-processing ensemble forecasts. Preliminary results from attempts to use the theory to speed the tuning of Hybrid data assimilation schemes will also be presented.

Assimilation of ozone data is an important and challenging component of atmospheric data analysis for operational and research purposes. In its operational analyses, the Global Modeling and Assimilation Office uses total column ozone data from the Ozone Monitoring Instrument (OMI) on the EOS Aura satellite and partial ozone columns from Solar Backscatter Ultraviolet (SBUV) sounders on NOAA platforms. These measurements are characterized by a relatively coarse vertical resolution of SBUV, particularly below the ozone peak, and coverage limited to the sunlit atmosphere. In this presentation we will discuss some recent changes introduced to ozone assimilation in the GEOS-5 Data Assimilation System (GEOS-DAS). We will describe a state-dependent approach to the background error covariance modeling and show how this approach leads to significant improvements in the representation of ozone in the Upper Troposphere and Lower Stratosphere. The second part of the talk will be devoted to assimilation of the Microwave Limb Sounder (MLS) data. Unlike the SBUV instruments, MLS can measure chemical composition and temperature in both day and night and possesses a much higher vertical resolution. We will show how MLS data can improve GEOS-DAS analyses and discuss our current work on direct assimilation of MLS radiance data in spectral bands sensitive to ozone and temperature.

A review of the history of the Joint Center for Satellite Data (JCSDA) is presented within the context of the ongoing successful transformation of the weather forecast process. Today, weather forecasts are becoming more accurate, with extreme weather events now predicted 4, 5, 6 and even 7+ days in advance in some cases. These improvements are driven in large part by improved numerical models, working off a global observing system, increasingly based on a wide range of satellite systems. The JCSDA represents a collaborative partnership among NASA, NOAA, Air Force and the Navy brought together to address the mission "to accelerate and improve the quantitative use of research and operational satellite data in weather, ocean, climate and environmental analysis and prediction models." The accomplishments of the JCSDA along with the current challenges/opportunities will be discussed, with an emphasis placed on the ongoing efforts to accelerate the transition of the new research and operational observing capabilities (advanced microwave, hyperspectral infrared, GPSRO, GOES-R) into the operational numerical prediction system. The presentation will conclude with a summary of the current prioritized efforts required to insure the rapid assessment and operational implementation of the JPSS and GOES-R sensors and upcoming research satellite missions as these systems come on line in the latter half of this decade.

The Suomi National Polar-orbiting Partnership (NPP) Advanced Technology Microwave Sounder (ATMS) provides 22 channels for probing atmospheric temperature and water vapor under all weather conditions. After intensive cal/val studies, the ATMS TDR data quality is similar to the Advanced Microwave Sounding Unit (AMSU) and Microwave Humidity Sounder (MHS). The global O-B distributions of ATMS temperature sounding channels are fairly uniform, which is a desirable characteristic for NWP applications. For quality control of clouds and precipitation affected radiances, the AMSU-A cloud liquid algorithm has been refined for ATMS applications. A new approach is developed for ATMS TDR to SDR conversion using the NPP pitch-over maneuver data. Using ATMS SDR data, the O-B displays less scan-angle dependence. On board the DMSP satellite, the Special Sensor Microwave Imager and Sounder (SSMIS) provides 24 channels for imaging and sounding the earth atmosphere. The radiance anomalies of the lower atmospheric sounding (LAS) channels were initially detected by analyzing NWP O-B and found to be associated with the radiation emitted by the antenna reflector and the solar contamination on the calibration targets. Algorithms have been developed to correct these anomalies. After the correction, impacts of SSMIS data on NWP medium-range forecast skills are demonstrated.

One of the many scientific challenges in the field of numerical weather prediction is to extract useful information on the atmospheric boundary layer using observations from remote sensing microwave instruments such as AMSU-A, AMSU-B/MHS or SSMIS. These data contribute increasingly to improve short to medium range forecasts, but also to improve re-analyses. Better use of these data often requires appropriate representation of the surface in the models, in both emissivity and temperature. This is achieved over sea, and satellite data have a tremendous impact on the atmospheric analyses over oceans. Over land, the surface emissivity is highly variable and may cause biases in the forward model if its temporal and spatial variability are not well taken into account. In such a situation, the model cannot produce realistic simulations of observations sensitive to the surface and may reject useful observations, including those not sensitive to the surface. This case concerns in particular the land and sea ice surfaces for which the surface emissivity is particularly challenging to model. During the seminar, some of the work carried out at Météo-France for a better description of the emissivity of land and sea ice surfaces will be summarized. The methodology used for estimating the emissivity in the model and its impact on the performance of the radiative transfer model will be presented. The impact of a proper modeling of the land and sea ice emissivity, in terms of forecast skill and quality of analyses, will be discussed.

The presentation will summarize some of the key updates of the ECMWF model and data assimilation systems over the past 2 years, after a brief introduction of ECMWF and its forecasting framework. Three examples of system upgrades will be given, namely the introduction of the ensemble of data assimilations, the revision of microwave sounder observation errors, and the new prognostic cloud scheme. Their respective impact on analysis and forecast skill will be demonstrated as well as the difficult trade-off between objective system improvements and positive impact on skill. Finally, the importance of observation monitoring in data assimilation will be highlighted with two examples.

To be provided.

The Suomi NPP satellite was successfully launched on October 28, 2011 and carries the following five sensors:

• Visible/Infrared Imager Radiometer Suite (VIIRS) providing advanced imaging and radiometric capabilities.
• Cross-track Infrared Sounder (CrIS) providing improved high vertical resolution atmospheric temperature and moisture information.
• Advanced Technology Microwave Sounder (ATMS) providing atmospheric temperature and moisture in all weather conditions.
• Ozone Mapping and Profiler Suite (OMPS) providing improved vertical and horizontal measurements of the distribution of ozone in the Earth's atmosphere.
• Clouds and the Earth's Radiant Energy System (CERES) sensor providing precise, calibrated global measurements of the earth's radiation budget

The Suomi NPP mission is the bridge between NOAA's Polar Operational Environmental Satellite (POES) and NASA's Earth Observing System satellites to the next-generation Joint Polar Satellite System, or JPSS, which NOAA will operate. NPP will provide on-orbit testing and validation of sensors, algorithms, ground-based operations, and data processing systems that will be used in the operational JPSS mission. The first JPSS spacecraft will be launched into the afternoon orbit by the middle of the decade to provide significantly improved operational capabilities and benefits; the last satellite in the JPSS series is expected to continue operations until about 2037. The JPSS program has an active program of user engagement to maximize the benefits of NPP and JPSS for critical products and services such as weather forecasting.

At the seminar, an overview of the JPSS program, some early results from each instrument, and user engagement will be presented.

The goal of the present study is to develop a method to assimilate Microwave Imager (MWI) brightness temperatures (TBs) into Cloud-Resolving Models (CRMs). To address the non-linear relationship of TBs to the state variables of CRM and the flow- dependency of the CRM forecast error covariance, we adopted an Ensemble-based variational data assimilation method (EnVA). In this presentation, I will report our recent studies on the following problems in EnVA:

1. Large-scale displacement errors of rainy areas between the observations and the CRM forecasts;
2. Serious sampling errors of cloud-physical variables because they were confined to rainy areas.

In order to solve the displacement error problem, we developed the EnVA that used Ensemble forecasts with displacement error correction. In order to alleviate the sampling error, we are introducing the following ideas to the EnVA:

1. Use of ensemble forecasts at neighboring grid points;
2. Classification of CRM variables and assumption of zero cross correlation between different classes.