A Research Plan for GOES-R Risk Reduction



Proposal to

National Oceanic and Atmospheric Administration

National Environmental Satellite, Data, and Information Service

Office of Systems Development

Washington DC

Research and Development for GOES-R Risk Reduction

Cooperative Institute for Research in the Atmosphere

Colorado State University

Foothills Campus

Fort Collins CO 80523-1375

Dr. Thomas H. Vonder Haar,

Principal Investigator

In partnership with

Dr. Mark DeMaria

Regional and Mesoscale Meteorology Team Leader

NESDIS Office of Research and Applications

Telephone: 970-491-8405

Email: Mark.DeMaria@

Debra Molenar

CIRA/Regional and Mesoscale Meteorology (RAMM) Team

Colorado State University

Telephone: 970-491-8447

Email: Molenar@cira.colostate.edu

July 2003

Period of Activity: July 1, 2003 – June 30, 2004

____________________________ __________________________

Dr. Thomas H. Vonder Haar, P.I. Mary Atella

Director, CIRA Senior Research Administrator

(970) 491-8448 Office of Sponsored Programs

(970) 491-2083

1. Introduction

The next generation GOES satellites (beginning with GOES-R) will include an Advanced Baseline Imager (ABI) and Hyperspectral Environmental Suite (HES) with vastly improved spectral, spatial and temporal resolution relative to the current GOES I-M series satellites. The GOES-R era will begin early in the next decade, and will be part of a global observing system that includes polar orbiting satellites with instruments with comparable spatial and spectral resolution. A three-year science study is proposed to use numerical simulations and existing in situ and satellite data to better understand the capabilities of these advanced instruments. This science study will help to reduce the time needed to fully utilize GOES-R as soon as possible after launch.

One of the advantages of GOES-R is the high temporal resolution that is possible from geostationary orbit. The emphasis of our science study is on three mesoscale atmospheric phenomena that evolve on time scales faster than that which can be sampled from polar orbiting satellites. Our study will focus on applications of GOES-R to tropical cyclones (with emphasis on intensity prediction), severe weather and mesoscale aspects of winter weather, including lake-effect snowfall. The approach is to use high-resolution numerical simulations coupled with radiative transfer models to determine sampling requirements, provide simulated data for development of retrieval algorithms and data assimilation strategies. In addition, we will assemble a multi-platform database of currently available earth-based and satellite (operational and experimental) observations for several cases of each type of weather system (tropical cyclone, severe weather and winter weather) to provide a sub-set of the type of data that will available from GOES-R. This database will be used for preliminary product development, to assess the utility of GOES-R instruments, and for training activities.

2. Atmospheric Phenomena

The Regional and Mesoscale Meteorology Team (RAMMT) located at CIRA has a long history of using satellite observations for research on severe weather, tropical cyclones, and mesoscale aspects of mid-latitude cyclones. GOES-R observations will be most applicable to these same atmospheric phenomena, due to the consistency between their physical time scales and the temporal sampling capability of the satellite.

Severe weather evolution is especially sensitive to the thermodynamic and dynamic environment surround the storms, and GOES-R will provide a unique opportunity to examine this sensitivity. Convection that develops into severe storms is often initiated by existing air mass boundaries on a variety of spatial scales, which can be monitored in the visible, near IR, and IR frequencies, since these areas are usually only partially cloudy before the convection develops. Thus, the HES may be especially useful to measure these structures, and to provide quantitative input for numerical forecast models. In addition, severe convection sometimes has unique cloud top structures such as over-shooting tops, and “enhanced-V” signatures that are often visible before severe weather is detected at the surface, or by Doppler radar.

The skill of tropical cyclone intensity forecasts is considerably less than the skill of track forecasts. Track forecasts primarily requires an accurate prediction of the synoptic-scale wind surrounding storm, but an intensity forecast requires an accurate estimate of the thermodynamic and dynamic properties of the storm environment, and the mesoscale processes with the storm itself. The GOES-R HES has tremendous potential for measuring the thermodynamic structure of the storm environment, and the potential to diagnose the inner structure for storms that have relatively cloud free eyes.

Although wintertime precipitation is often associated with synoptic-scale mid-latitude cyclones, there can be extreme local variations in snowfall amounts due to interaction with coastal fronts, gravity waves, and surface features such as the land/sea interface and large inland lakes. These local variations have fairly short time scales, and are often sensitive to the thermodynamic environments upstream from the precipitation. GOES-R data has the potential to improve the forecasts of local variations in winter precipitation through thermodynamic diagnosis from the HES, and the development of now-cast products from the ABI.

3. Case Study Database

Although observations similar to those from GOES-R might not become available prior to the GOES-R launch (depending on the outcome of the GIFTS program), “slices” of similar data are now or will become available well before GOES-R, from a variety of experimental and operational satellites and in-situ platforms. We propose to develop a multi-platform database for each of the three atmospheric phenomena described above. As a starting point, we will consider the following three cases:

1. Hurricane Lili (Sept. 21-Oct. 5, 2002). Lili formed in the Caribbean, rapidly intensified in the Gulf of Mexico, and then rapidly weakened just before making landfall along the Louisiana coast.

2. The May 7-10, 2003 Severe Weather Outbreak. During this four-day period, 1463 severe weather reports were logged by the NCEP Storm Prediction Center, including 223 tornados. This outbreak affected a large area from the front range of the Rockies through the Midwest.

3. The February 12-14, 2003 Lake-Effect Snow Storm “Newton”. A narrow band of snow formed downwind of Lake Ontario and dropped about 50 inches of snow centered on Oswego County, NY. A less intense snow band formed east of Lake Erie. This event was nicknamed “Newton” by the Buffalo NWS office, who informally names the most intense Lake-Effect snowstorms.

The pathway to HES includes the current GOES sounder, AIRS (2002), IASI (~2005), CrIS (~2006) and possibly GIFTS (2008?). For the initial case studies, current GOES sounder and AIRS data will be collected. The Hyperion data from EO-1 might also provide useful information concerning moisture variability. Although this data has a very narrow swath, it might also be useful for some limited areas. For imager applications, the data from GOES, AVHRR, AMSU, and MODIS will be collected initially. MSG data will also be collected for future tropical cyclone cases in the eastern Atlantic. It will also be possible to obtain TRMM data for the tropical cyclone studies. As the project evolves, data from NPOESS Prepatory Project (NPP) and other satellites will be included. In addition, in-situ data will be collected for ground truth and validation, including specialized in-situ observations such as hurricane aircraft reconnaissance and Kansas/Oklahoma ARM data when possible. The conventional rawinsondes, surface observations, etc will also be obtained for the case studies.

All of the this data will be converted to standard formats and used for science studies, advanced product development, input for numerical modeling studies, and training. A web site will be available for viewing the combined datasets. The longer-range goal of the project is to transition from case study to real time data displays, organized by atmospheric phenomena. For example, a “hurricane” site would provide multi-platform observations with common map backgrounds, enhancement curves, etc, to provide users with advanced looks at what will be available with GOES-R.

4. Numerical Modeling Studies

Perhaps the most difficult aspect of GOES-R to simulate using existing and pre-GOES-R satellite observations is the temporal resolution. For this reason, a numerical simulation component will be included in our science study. In the initial three-year research program, will we use numerical models in three ways, as follows.

1. Physical Process Studies - High resolution mesoscale models initialized will real data will be used to simulate each of the atmospheric phenomena described in section 2. The environments of simulated severe weather, tropical cyclones, and snow bands can be studied in detail to gain a better understanding of their temporal and spatial variability. For example, how rapidly does the near environment of a tornadic supercell or the temperature profile in the eye of a rapidly intensifying tropical cyclone vary with time, and how sensitive is the model evolution to this variation? These types of studies can provide justification of the need for thermodynamic sampling at a greater frequency than is possible from polar-orbiting platforms.

2. Simulated Satellite Observations - The numerical model output can be used as input to radiative transfer models to generate simulated satellite data. This simulated data can be used for advanced product development, and for input to satellite retrieval algorithms and data assimilation studies. This approach was particularly successful for the AIRS program, where simulated radiances were generated from the NCEP global model fields and radiative transfer models more than a year before the launch of the satellite.

3. Data Assimilation Studies – The simulated satellite data from the numerical model in combination with their data assimilation systems can be used to estimate the impact of various data types. For example, the initial condition from the model can be sub-sampled, and then used as input to a data assimilation system. It is then possible to address questions such as “How much would the forecast be improved if high resolution temperature profiles were available every ½ hour instead of every 4 hours, in a severe weather environment?” In addition, different data assimilation techniques can be tested to help determine optimal methods for using observations in various forms, and with variable accuracy and resolution. We will consider ensemble Kalman filtering assimilation techniques, which are being considered as the next generation method at NCEP.

For the numerical modeling studies, we propose to start with the Colorado State University Regional Atmospheric Modeling System (RAMS), which also includes an advanced data assimilation system. The RAMS model includes a comprehensive land/surface model, advanced microphysical parameterizations, and moving nested grids for hurricane simulations. This model is probably about five years more advanced than current operational mesoscale models used in NCEP operations. However, as the project progresses, we plan to make a transition to the Weather Research and Forecast (WRF) model. This model will likely become the next generation regional model at NOAA/NCEP, and also the next generation mesoscale model for the much of the research community. In addition, we plan to closely coordinate with the Joint Center for Satellite Data Assimilation (JCSDA), especially with regard to radiative transfer modeling. The JCSDA is developing a community radiative transfer modeling system, which we will incorporate into our research.

5. Project Personnel

Project Oversight will be provided by Prof. Thomas Vonder Haar and Ken Eis (CIRA/CSU) and Dr. Mark DeMaria (NESDIS/ORA).

The numerical modeling studies will primarily be performed by Drs. Milija Zupanski, Dusanka Zupanski, Louis Grasso, M. Sengupta with oversight from Mark DeMaria. Milija and Dusanka are CIRA employees who formerly worked for the NCEP Environmental Modeling Center, and are experts on mesoscale modeling and data assimilation. Louis Grasso has extensive experience with the RAMS model, and was a former student of Dr. Bill Cotton, the original developer of RAMS. M. Sengupta will lead the radiative transfer modeling activities. M. DeMaria is a NESDIS ORA employee with an extensive background in tropical cyclone analysis and forecasting, and numerical weather prediction.

The database development will be overseen by Debra Molenar and M. DeMaria. Debra leads the RAMM Team computer infrastructure group and has considerable experience in database management, satellite data processing and programming. A ¾ time CIRA computer programmer and ¾ time meteorologist will also be supported to assist with the database development, as well as with analysis of the numerical modeling output.

The project also will include a PhD student to focus on a specific aspect of the research, depending upon their interests, and a student hourly employee to assist with data processing. The inclusion of students will help cultivate future scientists that have familiarity with the GOES-R program.

6. Proposed Budget

The proposed budget for the first year of the project is $350 K. A detailed breakdown of the costs is provided in the attached spreadsheet. It is anticipated that the second and third year budgets will be lower (one the order of 300 K per year, depending on the availability of funds), because the first year includes $68 K for procurement of hardware for a mass storage device to house the extensive satellite database and numerical model output. Continuation proposals will be written for year 2 and year 3, at the discretion of NESDIS/OSD.

7. Project Coordination and Long Range Planning

It is anticipated that this research will be part of a larger GOES-R Risk Reduction program that will be coordinated with all of NESDIS/ORA. It is also anticipated that science activities will continue beyond the first three years proposed here. The continuation research will evolve towards making real time multi-platform datasets available for training and to familiarize the user community with potential GOES-R products. The numerical modeling studies will be generalized to assimilate real data (in addition to the simulated data proposed above), and to coordinate this effort with the on-going activities of the JCSDA. Prior to the launch of GOES-R, real-time simulated data should be produced on a routine basis in combination with a mesoscale model forecast at NCEP.

8. Progress Reports

Semi-annual progress reports will be provided to NESDIS/OSD management.

9. Budget Explanation

I and II. PERSONNEL and FRINGE BENEFITS

Salaries and benefits are requested for the personnel that will be performing this research, and providing administrative and computer support. In a basis consistent with our long-standing Memorandum of Understanding between NOAA and Colorado State University, the enclosed budget specifically includes support for administrative and clerical personnel (Fryer) directly associated with the technical and managerial administration of GIMPAP.  This support is “quid pro quo” for the reduced indirect cost rate agreed upon in the long-standing subject memoranda. K. Fryer will provide communication and collaboration support, assist in the acquisition and distribution of reference materials relevant to the conception and execution of the project, technical editing of scientific manuscripts, specialized reports and conference papers. All other budgeted personnel are directly involved in the research which is identified in the Statement of Work (SOW). 

III. DOMESTIC TRAVEL

Funds are requested for two trips to Washington, DC for coordination with ORA scientists. These trips are required to obtain the necessary radiative transfer modeling capabilities, and training on some of the specialized experimental satellite data to be used in this research. One trip to Madison is requested to attend the next NOAA Hyperspectral Satellite Data Workshop, planned for May of 2004.

IV. OTHER

1. The NT computer charges provide computer and data support associated with this project. The NT hourly rate is determined by CIRA and depends on the actual cost of the CIRA computer operations. CIRA charges Windows NT computer costs on an hourly basis, based on log-on time collected electronically via infrastructure programs.

2. Much of this research is performed in a McIDAS programming environment. A McIDAS Users Group (MUG) fee is necessary to use and update this software.

V. SUPPLIES

Storage media will be required to transport and back-up data and numerical model output associated with this project.

VI. TUITION

It is anticipated that one graduate student will be supported as part of this research project.

VII. EQUIPMENT

The hyperspectral data required for this project are extremely voluminous. Funds are included for a high-speed random access mass storage device to house these large satellite datasets.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download