Project title: Hurricane Forecast Improvement Project (hfip) – cira support to Tropical Cyclone Model Diagnostics and Product Development
Download 22.69 Kb.
|
|
PROJECT TITLE: Hurricane Forecast Improvement Project (HFIP) – CIRA Support to Tropical Cyclone Model Diagnostics and Product Development PRINCIPAL INVESTIGATOR: Wayne Schubert RESEARCH TEAM: Kate Musgrave, Scott Longmore, Andrea Schumacher, Louie Grasso, Robert DeMaria, Chris Slocum, Kathy Fryer NOAA TECHNICAL CONTACT: Philip Hoffman (NOAA/OAR) NOAA RESEARCH TEAM: John Knaff, Mark DeMaria (NOAA/NESDIS/STAR) PROJECT OBJECTIVES:
PROJECT ACCOMPLISHMENTS: Past Fiscal Year by Objective: 1—Retrospective forecasts using updated SPICE model -- One of the main accomplishments of the CIRA HFIP project has been the development of improved statistical intensity forecast models. One of these is called the Statistical Prediction of Intensity with a Consensus Ensemble (SPICE) model. It uses input from a number of dynamical models with a variety of tracks as input to two statistical models to form a consensus. Each hurricane season, the HFIP program runs experimental models in real time for evaluation by the National Hurricane Center. To become eligible for the demonstration, the model must be run on retrospective cases from the past three hurricane seasons. SPICE was designated for the real-time demonstration period in 2013 after completing the retrospective testing process. The SPICE model is currently undergoing upgrades and retrospective testing on the 2011-2013 hurricane seasons in preparation for the 2014 season. 2—Real-time forecasts using updated SPICE model -- The SPICE model was designated to run during the real-time demonstration period of the 2013 hurricane season due to its performance in the retrospective testing. SPICE was successfully run during that demonstration period from 1 August 2013 to 1 November 2013. Preliminary verification of the SPICE model for the 2013 hurricane season has been performed, with some of the results shown in Figure 1. SPICE had lower mean absolute errors in the Atlantic basin than either Decay-SHIPS or LGEM at longer forecast times.  
Figure 1. Mean absolute error (in kt) for forecast intensity for the 2013 Atlantic (left) and East Pacific (right) hurricane seasons for Decay-SHIPS (blue), LGEM (red), and SPICE (purple dashed). 3—Develop new hurricane forecast products: Environmental variable ensemble -- A program to diagnose a number of variables from hurricane models was developed under HFIP and is being improved by adding new parameters. Variables thought to be important for intensity change, such as vertical shear, relative humidity, etc., are being diagnosed from several real-time global and regional models, and products that show their differences are being developed and displayed in real time. A multi-model comparison plot featuring the operational models GFS, HWRF, and GFDL was provided to the HFIP products website (http://www.hfip.org/products/) during the real-time demonstration from 1 August 2013 to 1 November 2013 (Figure 2). 
Figure 2. Multi-model comparison plot for Hurricane Ingrid, valid at 0600 UTC 15 September 2013. This plot compares forecasts of TC and environmental variables from the three operational models GFS, HWRF, and GFDL. Clockwise from bottom-left: track (latitude and longitude of the center position), intensity (10 m maximum sustained winds), 850-200 mb vertical wind shear, sea surface temperature, and 700-500 mb relative humidity. Three additional models are included for the intensity forecasts: Decay-SHIPS, LGEM, and the HFIP experimental model SPICE. 4—Run experimental hybrid version of NHC’s operational windspeed probability program in 2013. Provide experimental products for HFIP products web page. Perform verification in post-season. -- An experimental version of NHC’s operational tropical cyclone wind speed probability model was run for the 2013 hurricane season, with graphical output provided to the HFIP products website (http://www.hfip.org/products/). The experimental version uses track information from a set of dynamical model ensembles, instead of from randomly sampling from the tropical cyclone track forecast errors from the past five years, as is used in the operational version. The intensity and structure perturbations are determined in the same way (using random sampling of forecast errors) as the operational version. The experimental version can represent more complex scenarios such as clustering of tracks and bimodal distributions. 5—Create an experimental version of SHIPS and LGEM, based on the ECMWF forecast fields. -- The operational versions of the statistical SHIPS and LGEM intensity models use input from the NCEP global forecast model. Work continues on adapting these models to use input from the ECMWF global forecast model, and the forecast errors will be compared with the operational version. The grib decoder routines used in the operational version of SHIPS and LGEM on NCEP's WCOSS system have been modified to use 1 degree ECMWF model output instead of the 1 degree GFS model output. A parallel version of the operational SHIPS/LGEM script was also created with a switch to run off the ECMWF fields instead of the GFS fields. A side benefit of this approach is that the operational Rapid Intensification Index (RII) will also be run off the ECMWF fields for comparison with the GFS version. Initial tests showed that the ECMWF grib files on WCOSS did not contain all the fields needed to run SHIPS and LGEM. When more complete ECMWF fields are provided, the development of the experimental SHIPS, LGEM and the RII will continue. The ECMWF versions of SHIPS and LGEM may be candidates for inclusion of future versions of the SPICE model. 6—Collect observed GOES data for 2013 storm cases. Provide data to other groups, and perform verification of HWRF synthetic imagery. Add satellite total precipitable water for evaluation of model moisture fields. -- One of the difficulties of verifying hurricane models is the lack of observations near the storm, especially in the upper levels. Observations of the moisture fields are also very sparse near tropical cyclones. To aid in the evaluation and verification of the HFIP forecast models, synthetic GOES satellite data and total precipitable water (TPW) fields from the model output are being compared to the real GOES satellite imagery and a microwave-based satellite TPW product. This comparison will help to identify error sources in the models, and provide feedback to users on areas where the models can be improved. To facilitate this comparison observed GOES data and satellite precipitable water was collected for the 2013 hurricane season. Verification of HWRF synthetic imagery for the 2013 season has successfully been performed, with some of the results presented in Figure 3. The current operational HWRF shows reduced biases, particularly in the synthetic IR brightness temperatures, as compared to the previous version of HWRF. Satellite total precipitable water has been added, and Figure 4 shows a comparison of imagery generated from HWRF total-column precipitable water and blended satellite total precipitable water.  
Figure 3. Mean absolute error (solid) and bias (dashed) for the HWRF synthetic Channel 3 (WV, blue) and Channel 4 (IR, red). The left panel highlights the period of the 2013 Atlantic hurricane season before the current operational version was implemented; the right panel shows the operational HWRF for the period from August 2013 through the end of the Atlantic hurricane season. 
Figure 4. HWRF total-column precipitable water (left) and blended satellite total precipitable water (right) for Tropical Storm Erin, 12 UTC 18 August 2013. PUBLICATIONS refereed: Hurricane Forecast Improvement Project (HFIP) – CIRA Support to Tropical Cyclone Model Diagnostics and Product Development DeMaria, M., J.A. Knaff, M. Brennan, D. Brown, R. Knabb, R.T DeMaria, A.B. Schumacher, C. Lauer, D. Roberts, C. Sampson, P. Santos, D. Sharp, and K. Winters, 2013: Improvements to the operational tropical cyclone wind speed probability model. Wea. Forecasting, 28, 586-602. doi: http://dx.doi.org/10.1175/WAF-D-12-00116.1 DeMaria M, J.A. Knaff , R. Zehr, 2013: Assessing hurricane intensity using satellites. Satellite-based applications to climate change. J.J. Qu, A. Powell, and M.V.K. Sivakumar, Eds, Springer, New York, pp 151-163. doi: http://dx.doi.org/10.1007/978-94-007-5872-8_10 DeMaria, M., C.R. Sampson, J.A. Knaff, and K.D. Musgrave, 2014: Is tropical cyclone intensity guidance improving? Bull. Amer. Meteor. Soc., in press. Knaff, J.A., M. DeMaria, C.R. Sampson, J.E. Peak, J. Cummings, W.H. Schubert, 2013: Upper oceanic energy response to tropical cyclone passage. J. Climate, 26, 2631-2650. doi: http://dx.doi.org/10.1175/JCLI-D-12-00038.1 Knapp, K.R., J.A. Knaff, C. Sampson, G. Riggio, and A.D. Schnapp, 2013: A pressure-based analysis of the historical western North Pacific tropical cyclone intensity record, Mon. Wea. Rev., 141, 2611-2631. Lin, I-I, G.J. Goni, J.A. Knaff, C. Forbes, M.M. Ali, 2013: Tropical cyclone heat potential for tropical cyclone intensity forecasting and its impact on storm surge. Journal of Natural Hazards. 66,1481-1500. doi:10.1007/s11069-012-0214-5 Slocum, C. J., G. J. Wililams, R. K. Taft, and W. H. Schubert, 2014: Tropical cyclone boundary layer shocks. J. Adv. Model. Earth Syst., accepted. Zhang, Man, Milija Zupanski, Min-Jeong Kim, John A. Knaff, 2013: Assimilating AMSU-A Radiances in the TC Core Area with NOAA Operational HWRF (2011) and a Hybrid Data Assimilation System: Danielle (2010). Mon. Wea. Rev., 141, 3889–3907. doi: http://dx.doi.org/10.1175/MWR-D-12-00340.1 PRESENTATIONS: Hurricane Forecast Improvement Project (HFIP) – CIRA Support to Tropical Cyclone Model Diagnostics and Product Development DeMaria, M., A. Schumacher, and K. Musgrave, 2013: A reformulation of the logistic growth equation model (LGEM) for ensemble and extended range intensity prediction. HFIP conference call, August 7. DeMaria, M., J. Knaff, K. Musgrave, A. Schumacher, R. DeMaria, L. Grasso, S. Longmore, and C. Slocum, 2013: 2013 NESDIS HFIP activities. HFIP conference call, October 23. DeMaria, M., K. Musgrave, A. Schumacher, L. Grasso, J. Knaff, and D. Lindsey, 2013: NESDIS/CIRA Diagnostics - 2013. 2013 HFIP Diagnostics Workshop, November 22, College Park, MD. Slocum, C. J., K. D. Musgrave, L. D. Grasso, G. Chirokova, M. DeMaria, and J.A. Knaff, 2013: Satellite applications to hurricane intensity forecasting. CoRP Science Symposium, July 23-24, Madison, WI. Directory: ftp -> Borger -> CIRA-RAMMB-AnnualReports March2014 ftp -> Federal Communications Commission fcc 12-81 Before the Federal Communications Commission ftp -> Tr-41. 4-03-05-024 Telecommunications ftp -> Message Race Detection for Web Services by an smt-based Analysis ftp -> A proposal submitted to the ftp -> National reports on data buoy activities ftp -> Louisiana barrier island comprehensive monitoring program (bicm) Volume 1: Barrier Shoreline Post-Storm Assessment ftp -> Jcomm expert team on CIRA-RAMMB-AnnualReports March2014 -> Project title: Feature-based Validation of mirs soundings for Tropical Cyclone Analysis and Forecasting CIRA-RAMMB-AnnualReports March2014 -> Principal investigator: Renate Brummer research team Download 22.69 Kb. Share with your friends:
|