J3SCOInc. Applied Meteorology Unit ____ (AMU) Quarterly Report 1980 N. Atlantic Ave., Suite 230 Coca Beach, FL 32931 (321) 783-9735, (321) 853-8202 (AMU) First Quarter FY-05 Contract NAS1O-Ol 052 31 January 2005 Executive Summary NASA HO/M/F. Gregory NASA KSCLAAIJ Kennedy This report summarizes the Applied Meteorology Unit (A MU) activities for the first quarter of Fiscal Year NASA KSC/MK/D KrOSS 2005 (October - December 2005). A detailed project schedule is included in the Appendix. NASA KSC/PI-i/ M. Viletmore NASA KSC/PH-A2/D Lyons NASA KSC/PH/M. Leinbach Task Objective Lightning Probability Forecast: Phase I NASA KSC/PHIS Mmute NASA KSC/PH-A/J Guidi Goal Develop a set of statistical equations to forecast the probability of NASA KSC/YNJ Heaki lightning occurrence for the day. This will aid forecasters in evaluating NASA KSC/YA-C/O Bartine NASA KSC/YA-D/G Clements flight rules and determining the probability of launch commit criteria NASA KSC/YA-DIJ Madura violations, as well as preparing forecasts for ground operations. NASA KSCYAD/F Merceret NASA JSC/MAIW. Parsons NASA JSC!MS2/C, Bin Milestones A graphical user interface (GUI) was developed to allow forecasters NASA JSCIZS8/F. Brody user-friendly access to the equations. The GUI was developed through NASA JSCIZSB/R. Lafosse consultations with the 45th Weather Squadron (45 WS) to ensure the NASA JSC/ZS8/B Hoeth NASA MSFC/ED41IIN Vaughan final product meets their operational needs. NASA MSFC/EV13IB Roberts NASA MSFC/EV13/S. Deaton Discussion The GUI allows the user to input values that are readily available to NASA MSFC/MP71/G Overby NASA DFRCIR.AJJ Ehemberger operational forecasters from the morning sounding, then calculates the 45WS/CC/R LaFebre probability of lightning occurrence for the day and displays the value 45 WS/OOfP Broil quickly in an easy-to-read format. 45WS/DOR/F FUnn 45 WSIOORJT. McNamara 45WSIDORTT. Longrnire Task Severe Weather Forecast Decision Aid 45 WS/DORIJ. Tumbiolo 45WSIDOR/K Wnters Goal Create a new forecast aid to improve the severe weather watches and 45 WS/DORIJ. Weerns 4SWS/SY/K Beilue warnings for the protection of Kennedy Space Center (KSC)/Cape 45WS/SYPJB Boyd Canaveral Air Force Station (CCAFS) personnel and property. 45 WS/SYR/W. Roeder 45 RMS/CC/A Lester Milestones The analysis to separate lightning vs. non-lightning days was 45 RMS/LGP/R. Fore 45 SWISESLID. Berlinrut completed. The stability indices were stratified by severe weather, 4SSW/XPR!R Hiflyer lightning but no severe weather, and no lightning days, then analyzed to 45 OG/CC/G, Eu/nan CSR 4500/H. Hernng determine if there were significant differences in their values between CSR7000/M Mauer severe and non-severe weather on lightning days only. SMC/RNP1D. Satm SMC/RNP/T Knox Discussion While the analysis of the stability indices is not complete, initial results SMC/RNP/R Bailey SMC/RNP (PRC)/P Conant still show that no significant differences in values exist that will help HO AFSPC/XOSW/A Gibbs forecasters discern between a severe/non-severe weather day. HO AFWNDNXJW Cade HO AFWNDN/M. Zetttemoyer Task Hail Index HOARNNON/M. Sumreier HQ AFWA/XORJM Tree HO USAF/XOW/R Clayton Goal Evaluate current techniques used by the 45 WS to forecast the HO USAF/XOWX/H Elkins probability of hail occurrence and size. Hail forecasts are required to NOAA 'W/NP"IL. Uccellunu NOAA/OAR/SSMC-I/J. Golden protect personnel and material assets at KSC, CCAFS, Patrick Air Force NOAAJNWS/0ST12/SSMC2JJ. McQueen Base, and the Melbourne International Airport. The evaluation results NOAA Off ice of Military Affairs/N. Wyse NWSMelbourne/B Hagemeyer will be used by the 45 WS to determine if a new technique is needed. NWS Southern Region HQ/'W/SRH/ X. W Proonza Milestones The AMU obtained a copy of the software used to generate the current NWS Southern Region Harw/sR3 hail size forecasts from sounding data. The computed forecasts were O Smith NWSPW/OST1/B. Saft/e compared to observations, revealing a large discrepancy in the values NWSrwIOST1 2/D. Melendez forecast vs. those observed. Two major errors were found in the NSSL/D. Forsyth NSSUC. Crisp software and reported to Computer Sciences Raytheon personnel. Continued on Page 2 Discussion The errors in the code were modified in the AMU copy of the software and new forecasts were computed. The results showed a closer association between the number of hail events forecast and the number ______________________ observed. . ... - Fage.2,,. 'T!.!J Distribution (cceteued from Page 1) Executive Summary, eoiitillu!e1I 30 WS!SY/M Sdirneiser 3CWS/SYR/L Wells 3OWS/CCIC. Keith 30 WS/DO/M Fitzgerald Task Stable Low Cloud Evaluation 30 SW!XPE/R. Ruecker 88 WSIWESIK Lehneis Goal Examine archived data collected during rapid stable cloud development 88WSIWES/G Marx 46 WSIIDO/J. Mackey events resulting in cloud ceilings below 8,000 ft at the Shuttle Landing 46 WSIWST/k Midiels Facility (SLF). Document the atmospheric conditions favoring this type 412 OSS/OSWM/P. Harvey of cloud development to improve the ceiling forecast issued by the FSU Department of Meteorology/H Fuelberg Spaceflight Meteorology Group (SMG) for Shuttle landings at KSC. ERAUlApplied Aviation Sciences/ C.Herbster Milestones Used local rawinsonde and hourly surface observations at the SLF to ERAU/CAAR/L Wilson NCARJJ Wilson develop a database of days with low-level temperature inversions and NCARTY H. Kuo cloud ceilings below 8,000 ft using data from 10 cool seasons NOMJERL/FSUJ. McGinley Office of the Federal Coordinator for (November-March) from the years 1993-2003. Meteorological Services and Supporting Research/J Harrison. R. Dumont Discussion An objective analysis was done first by using software to choose days 8oeing HoustordS. Gonzalez Aerospace Corp/i. Adarig with low-level inversions and days with ceilings below 8000 ft. A ACTA Inc lB. Parks subjective analysis was then performed to match days that had both ENSCO, lnc./T. Wilfong ENSCO. Inc/B Larnbert attributes and moisture near the inversion level. These days will be FNRCO len /A fliann compared to satellite data to build a database of rapidly developing ceiling events. Task Shuttle Ascent Camera Cloud Obstruction Forecast Goal In response to a Shuttle Program request to implement a Columbia Accident Investigation Board recommendation, develop a model to forecast the probability that at least three of the shuttle ascent imaging cameras will have a view of the shuttle launch vehicle unobstructed by cloud at any time from launch to Solid Rocket Booster separation. Milestones Completed the customer-recommended revisions to the final report. The AMU remains on standby to present briefings to the Shuttle Launch Director and Integration Control Boards as required. Discussion Results from this work were presented at the American Meteorological Society's 11th Conference on Aerospace, Range, and Aviation Meteorology. Task RSA and Legacy Wind Sensor Evaluation Goal Compare wind speed and direction statistics from legacy and RSA sensors on the Eastern and Western Ranges (ER and WR) to quantify the impact of sensor technology changes on wind measurements. The 45 WS and 30th Weather Squadron (30 WS) use these winds to issue weather advisories for operations and, therefore, need to know if there are significant differences in the wind measurements between the systems. Milestones Began archiving ER legacy data and received samples of ER RSA sonic wind data for preliminary analysis. Software to ingest and reformat the data was developed and tested. Discussion A preliminary analysis of the 1-second wind speed data from the ER RSA sensors on Tower 313 did not show the degree of variability expected. The data capture protocol and data transfer interfaces are being examined to answer questions resulting from the preliminary analysis. Continued on Page 3 - __________ U Executive Summary, conlirnieti TABLE of CONTENTS Task ARPS Optimization and Training Extension SHORT- TERM FORECAST Goal Provide assistance and support for upgrading and improving the IMPROVEMENT operational Advanced Regional Prediction System (ARPS) and ARPS Data Analysis System (ADAS) at the National Weather Service (NWS) Objective Lightning in Melbourne (MLB) and SMG forecast offices. Probability.......................4 Milestones Completed the upgrade of the ARPS numerical weather prediction Severe Weather Forecast model to software version 5.1.2 at the NWS MLB. Decision Aid....................6 Stable Low Cloud............8 Discussion The new version of ARPS has many improvements and features that will make maintenance much easier. The ADAS upgrade requires Hail Index........................9 additional work to install the necessary supportive software components Shuttle Ascent Camera on the new Linux system. Cloud Obstruction Forecast Task User Control Interface for ADAS Data Ingest ....................11 Goal Develop a GUI to help forecasters manage the data sets assimilated INS TRUMEN TA TION AND into the operational ADAS run at NWS MLB and SMG. MEASUREMENT Milestones Added two components to the GUI that allow the user to determine l&M and RSA Support... 11 which groups of observations are analyzed by ADAS, and have the RSA and Legacy Wind ability to modify the display of observation data logs. Sensor Evaluation.........11 Qiscussion One new feature gives the user power to choose which groups of MESOSCALE MODELING observations are used in the ADAS analysis by turning them on or off. The second feature allows the user to see a data log of the number of ARPS Optimization and observations being ingested into ADAS at a particular time or a flag Training Extension ........12 indicating whether a data type was used. User Control Interface for ADAS Data Ingest.........12 AMU CHIEF'S TECHNICAL A C Ti Vi TIES .....................16 AMU OPERATIONS.....16 REFERENCES ............. 16 LIST OF ACRONYMS.. 17 APPENDIXA................ 18 Special Notice to Readers Applied Meteorology Unit (AMU) Quarterly Reports are now available on the Wide World Web (WWW) at http://science. ksc. nasa.gov/amu/. The AMU Quarterly Reports are also available in electronic format via email. If you would like to be added to the email distribution list, please contact Ms. Winifred Lambert (321-853-8130, [email protected]). If your mailing information changes or if you would like to be removed from the distribution list, please notify Ms. Lambert or Dr. Francis Merceret (321-867-081 8, Francis.J. Merceret©nasa.gov). BackgrouNd The AMU has been in operation since September 1991. Tasking is determined annually with reviews at least semi-annually. The progress being made in each task is discussed in this report with the primary AMU point of contact reflected on each task. AMU ACCOMPLISHMENTS DURING THE PAST QUARTER SHORT- TERM FORECAST The results of the equation development and IMPRO VEMENT testing were shown in the previous AMU Quarterly Report (Q4 FY04). All of the equations showed an increase in skill over the benchmark forecasts of Objective Lightning Probability daily and monthly climatology, persistence, and (Ms. Lam bert and Mr. Wheeler) the flow regime lightning probabilities. As a result, the new equations will be added to the current set The 45th Weather Squadron (45 WS) of tools and procedures used by the 45 WS forecasters include a probability of thunderstorm forecasters to make the daily lightning probability occurrence in their daily morning briefings. This forecast. information is used by personnel involved in determining the possibility of violating Launch Graphical User Interface Commit Criteria (LCC), evaluating Flight Rules, and planning for daily ground operation activities In order to use these equations, the on Kennedy Space Center (KSC)/Cape Canaveral forecasters need an interface that will facilitate Air Force Station (CCAFS). Much of the current user-friendly input and fast output. In cooperation lightning probability forecast is based on a with Mr. Roeder of the 45 WS, Ms. Lambert subjective analysis of model and observational developed a graghical user interface (GUI) using data. The forecasters requested that a lightning Microsoft® Excel Visual Basic®. The 45 WS was probability forecast tool based on statistical involved in the GUI development by providing analysis of historical warm-season data be comments and suggestions on the design. This developed. Such a tool would increase the will ensure that the final product will address their operational needs. objectivity of the daily thunderstorm probability forecast. The AMU developed statistical lightning The GUI begins with a dialog box that queries forecast equations that provide a lightning the user for the current month and day, shown in occurrence probability for the day by 1100 UTC Figure 1. A drop-down list is shown for each (0700 Eastern Daylight Time (EDT)) during the parameter by clicking on the arrow to the right of months May - September (warm season). The the text box. The user must choose a value from tool is based on the results from several research the Month drop-down list, but has the option of projects. entering a day value manually or through the Day drop-down list. Choosing the month determines J!1 which equation will be used, as there is one "Calculate Another Probability" button, this dialog equation for each month. Choosing the day will box is closed and control is returned to the second determine the Daily Climatology value used as a dialog box containing the predictor choices. At this predictor in the equation. point the user can make new choices for the predictors and calculate a new probability, or click I083EtT1 UQITNINt FORECA . :41 the "New Date" button and return control to the first dialog box. At this point the user can choose Today's Date a new date and start the process over, or click the Cancel" button and close the GUI. Month Jul j Choose the Month, _________ then the Day from the Day J 15 :] drop-down lists Persistence Was iihiing observed i at least one of the KSC/CCAFS advisory circles yestarday Cancel I F Continue.,. No between 0700 - 2400 EDT? ___J Flow Regime Figure 1. The first dialog box in the GUI that SW: Low-level (1000-700 mb)ridge South of XMR queries the user for the month and day. Month (SW-i and SW-2 regimes corribined) and Day are chosen by clicking on the down arrows and choosing from the drop-down SE Low-level ndge North of XMR (SE-i and SE-2 regimes combined) lists. The Cancel button exits from the GUI, the Continue button brings up the next dialog box. C Uniform NW flow across the peninsula Once the Month and Day are chosen, the user C Uniform NE flow across the peninsula clicks on the "Continue.....button. This executes a second dialog box that contains choices for C Other: None of the above predictor values based on the month. The dialog box in Figure 2 contains the predictor choices for Obtain the fob wing data values from the July, following the choices in Figure 1. The MILOS kew-r product: predictors for the daily lighting probability in July Total Totals (TT) are 1-day persistence, flow regime, Total Totals Entar the Total Totals from this (TT), and the average relative humidity in the 800- j morning's 1000 Z XMR sounding 600 mb layer (RH). The GUI allows the user to choose only one item under Persistence and only Average 800-600mb RH one item under Flow Regime. The values for TI and RH from the morning 1000 UTC sounding can Enter the average 800 - 600 mb layer relative humidity from this nu-ning's be input manually or by using the up/down arrows 62J j 1000 Z XMR soundrig (rounded to the right of the text box. The initial values for iteger value without %, e.g. enter the sounding parameters in all months (TI and 65.2% as 65, 65.7% as 66) RH in July) are set to -999 when the predictor dialog box is first displayed. If the user forgets to input values for any of the sounding parameters, New Date Calculate Probability.,. the -999 value will trigger an error dialog box that tells the user to input a value for the parameter. Figure 2. The second dialog box that contains The GUI code contains upper and lower limits on choices for the predictors in the July equation. the parameter values to ensure realistic values Persistence and Flow Regime are chosen by are entered. clicking one of the option buttons in each After making all the choices in the predictor section. IT and RH are chosen by entering the dialog box, the user clicks on the "Calculate values manually or using the up/down arrows. Probability.....button. This executes the third and The "New Date" button closes this dialog box final dialog box that displays the lightning and returns control to the first dialog box probability for the day as calculated by the (Figure 1). The "Calculate Probability..." equation for the month. Figure 3 shows the button brings up the next dialog box. lightning probability for the values shown in Figures 1 and 2. When the user clicks the 'F J ThT - For more information on this work, contact O8ABIUTY OF Ms. Lambert at lambert.wintfredensco.com or 321-853-8130. The probability of lightning being observed in at least one of the Severe Weather Forecast Decision Aid KSC/CCAFS advisory circles today (Mr. Wheeler and Dr. Bau man) from 0700 - 2400 EDT is: The 45 WS Commander's morning weather briefing includes an assessment of the likelihood of local convective severe weather for the day in 84% order to enhance protection of personnel and material assets of the 45th Space Wing, CCAFS, and KSC. The severe weather elements produced by thunderstorms include tornadoes, wind gusts ^ 50 kts, and/or hail with a diameter ^ 0.75 in. Forecasting the occurrence and timing of these CaIculat Another Probability phenomena is challenging for 45 WS operational personnel. The AMU has been tasked with the creation of a new severe weather forecast decision aid, such as a flow chart or nomogram, to Figure 3. The third and final dialog box in the improve the various 45 WS severe weather GUI containing the probability of lighting for watches and warnings. The tool will provide the day based on the inputs shown in Figures severe weather guidance for the day by 1100 I and 2. UTC (0700 EDT). Excel Workbook Dr. Bauman created a lightning strike The GUI described in the previous section is database using Cloud-to-Ground Lightning executed from an Excel° workbook containing six Surveillance System (CGLSS) data for May- worksheets. The first worksheet contains brief September 1989-2003. The CGLSS data was instructions on how to start and use the GUI, and obtained from Mr. Paul Wahner of CSR in the the other five pages contain information for each form of text files containing the date, time, latitude, individual month. The information for each month longitude, strength, and polarity of individual includes the cloud-to-ground lightning strikes. The ArcView® • Predictor names and their coefficients in the Graphical Information System (GIS) software equations, package was used to display the data on a map background in order to confirm the existence of • Flow regime names and their probabilities thunderstorms in the area of interest. In order to of lightning occurrence, import the data into the GIS package, they were • Climatological lightning probability for each reformatted into a single file for each day, then day, converted to ExceiC files, and finally saved in • Minimum, maximum, median, mean, and dBase IV format files. The displays were used to first and third quartile values of the distinguish between thunderstorm and non- observed sounding stability indices, thunderstorm days. A previous analysis using stability indices from all days was unable to • Range of valid values in the GUI for the produce any correlations between index values stability indices, and and severe weather. Dr. Bauman and Mr. • Stability index values associated with Wheeler thought an analysis using only days with convection. lightning may find a distinction between severe and non-severe thunderstorm days. After choosing a month and day in the first dialog box and continuing, the worksheet The database of all stability indices was corresponding to the chosen month is displayed in stratified into three categories: the workbook along with the second GUI. This • Severe weather days, allows the user to view all the possible parameter values for use in a particular month's equation. • Lightning but no severe weather days, and • Non-lightning days. L JP There are three important caveats to note in IT values for each day-stratification in Figures 4 this stratification. First, a day with no severe through 6. For comparison of values between the weather does not mean severe weather did not three stratifications, Figure 7 shows the IT for all occur in the area of interest, but that no severe days combined. weather was reported. There is no other way to verify severe weather. Second, a non-lightning In each figure there are two horizontal lines day only means there was no cloud-to-ground showing the current 45 WS values for lightning. There could have been in-cloud or distinguishing between low, medium and high risk cloud-to-cloud lightning that the CGLSS sensors for severe weather. Figure 4 shows the U values cannot detect. as red dots on days in which severe weather was reported. The physical expectation was that the The 14 stability parameters listed below were dots would be clustered toward high U values examined in each of the three categories: and in or near the risk category. in fact, the scatter • Total Totals (IT), in the plot is large with U covering the range of values. The same analysis can be made for • K-Index, Figure 5, which shows the U values as green • Lifted Index, dots on days with lightning but no severe weather. • Thompson Index, The analysis for non-lightning days, shown by • SWEAT Index, blue dots in Figure 6, is quite different. Most of the values are clustered in the low risk category. • Precipitable water, However, when all values are combined in the • Showalter Stability Index, scatter plot shown in Figure 7, one can see that a • Cross Totals, low U value does not necessarily reduce the risk of lightning or severe weather occurrence. • Convective Available Potential Energy (CAPE), it is clear from the U examples that this • CAPE Max Oe (CAPE based on layer with stability parameter is not a good indicator to discern among the three types of days. The other maximum equivalent potential temperature), stability parameters do not appear to be good • CAPE FMaxT (CAPE based on forecast objective indicators of severe weather occurrence maximum surface temperature), but they are being studied in detail. In addition to • Convective Inhibition (CIN), testing the feasibility of using the individual • T500 (500 mb temperature), and stability parameters as stand-alone discriminators, other criteria such as surface high pressure ridge • T (500 mb dew point temperature). d500 position, persistence, and upper level dynamics Scatter plots of each stability parameter were are being studied individually and in combination created for each individual day-stratification and with the stability parameters. for all days combined. These scatter plots for Finally, Dr. Bauman and Mr. Wheeler each stability parameter were used to try and continued working on their proposed Severe discern if the morning CCAFS sounding was a Weather Decision Aid, essentially an update to good tool to help objectively differentiate among the current 45 WS severe weather checklist, and reported severe weather days, thunderstorm days began writing the final report. with no reported severe weather, and days with Contact Mr. Wheeler at 321-853-8205 or no lightning. The ultimate goal was to determine if wheeler. mark(ensco.com, or Dr. Bauman at 321- a signal was present in any of the stability 853-8202 or bauman.billensco.com for more parameters to help the 45 WS forecaster analyze information on this work. the thunderstorm and severe weather threat. Examples of the scatter plots are shown using the Total Totals Total Totals Severe Weather Days Lightning Days May- September 1989-2003 May - September 1989-2003 55 55 High - 50 50 ,_, _--..:. .. :....-.. .. ::. 45 15 '-' 1 '0 ' - 40 .5 - LeO 40 35 0 500 1000 5500 2000 35- Ras,nuonde Number 0 500 1000 1500 2000 Rawineonde Number Figure 4. Scatter plot of TT (red dots) for all Figure 5. Scatter plot of TT (green dots) for all days with reported severe weather. The days with lightning but no severe weather. The rawinsonde number is shown along the x-axis rawinsonde number is shown along the x-axis and represents soundings in the months May— and represents soundings in the months May— September, 1989-2003. The heavy black lines September, 1989-2003. The heavy black lines sub-divide the severe weather threat sub-divide the severe weather threat categories of low, medium, and high IT values categories of low, medium, and high IT values based on the current 45 WS severe weather based on the current 45 WS severe weather checklist. checklist. Total Totals Total Totals Non-Lightning Days Severe Weather, Lightning, and Non-Lightning Days May - September 1989-2003 May -September 1989-2003 00 - Liysning Dayc No,-Srwn . Sr,n,e . Nw. 9,1r.ng Oatw 51 High High 50 - Hed5m I -. C -- —: 40 .. .. . • . . .. •.. 0 --'.: 'e.2.5 LoN .•. . . .. ...... .. 35 0 500 1000 1500 2000 0 500 1000 1500 Rewinsonde Number I Rawinsende Number Figure 6. Scatter plot of IT (blue dots) for all Figure 7. Combination of the U scatter plots days with no lightning. The rawinsonde for the three stratifications. The rawinsonde number is shown along the x-axis and number is shown along the x-axis and represents soundings in the months May— represents soundings in the months May— September, 1989-2003. The heavy black lines September, 1989-2003. The heavy black lines sub-divide the severe weather threat sub-divide the severe weather threat categories of low, medium, and high IT values categories of low, medium, and high U values based on the current 45 WS severe weather based on the current 45 WS severe weather checklist. checklist. Stable Low Cloud Evaluation are the most challenging to predict accurately. (Mr. Wheeler and Mr. Case) The AMU is tasked to develop a database of these cases, identify the onset, location, and if Forecasters at the Space Meteorology Group possible, dissipation times, and document the (SMG) issue 30 to 90 minute forecasts for low atmospheric regimes favoring this type of cloud cloud ceilings at the Shuttle Landing Facility (SLF) development. for all Space Shuttle missions. Mission verification The forecasters at SMG indicated that these statistics have shown cloud ceilings to be the events typically take place in the cool season biggest forecast challenge. Forecasters at SMG during daylight hours. Therefore, Mr. Case and are especially concerned with rapidly developing clouds/ceilings below 8000 ft in a stable, capped Mr. Wheeler collected the morning CCAFS rawinsondes (XMR) and hourly surface thermodynamic environment, since these events :4.' 1 observations at the SLF (US) between 1100- used to make daily hail forecasts and, if needed, 2300 UTC during the cool season months of to develop a new tool tuned to the local area. November-March 1993-2003, for a total of 10 cool Dr. Short and Mr. Wheeler evaluated the seasons. Mr. Case developed software to identify operational technique used by the 45 WS to inversions below 8000 ft in the XMR data and forecast hail. The Neumann-Pfeffer index ceilings less than 8000 ft in the US surface (Neumann 1971) for forecasting thunderstorm observations. The software provided daily output probability is combined with the Fawbush-Miller of the low-level inversions and cloud ceilings, as hail graph (Reymann et al. 1998) for forecasting well as reports of precipitation and fog at US for hail size. Dr. Short and Mr. Wheeler generated all 10 cool seasons. hail forecasts for the 14-year period 1989-2002 Mr. Wheeler then conducted a subjective using an archive of XMR rawinsonde data and the analysis of the output in order to identify potential operational computer programs run by CSR case days. Through this subjective analysis, the personnel at the CCAFS weather station. Results database was narrowed further to exclude of the comparison between hail size forecasts and precipitation events and days with ceilings below hail size reports in Brevard County, Florida are 8000 ft all day, since the goal of the task is to presented below. study the ceiling development. Days were Climatology of Forecast Hall Sizes from the 45 identified as potential events if they exhibited each WS "Thunderstorm Probability Study" of the following three elements: The probability of thunderstorm activity for the • A low level inversion, day and the expected size of associated hail is • Moisture near the inversion, and determined each morning by 45 WS forecasters • A ceiling below 8000 ft. through analyses of the local and large-scale weather patterns as well as data from rawinsonde Mr. Wheeler and Mr. Case will obtain and observations (RAOBs) taken at XMR. The CSR examine satellite data for the potential case days personnel at XMR use a computer program to determine the actual events for which there named WVTHUV1 .f to analyze vertical profiles of were rapidly developing ceilings below 8000 ft. temperature, humidity and wind from the RAOBs One of the requirements of this task is to examine to generate a bulletin titled "Thunderstorm only events when the low ceilings occurred as a Probability Study." The bulletin includes a yes/no result of development or formation, not advection indicator for thunderstorm activity and a forecast from another location. The satellite imagery will be hail size. The bulletin is transmitted to the important in determining whether the low cloud Meteorological Interactive Data Display System ceilings were caused by formation or advection. (MIDDS) in the Weather Operations Center and The final database will contain only days with low can be displayed by entering the MIDDS cloud ceiling development, excluding days when command "CYA MISC THUNDER". The clouds advected over central Florida. Thunderstorm Probability Study bulletin is Contact Mr. Wheeler at 321-853-8205 or generated during the warm season months of wheeler.markäensco.com, or Mr. Case at 321- May-September. The areas of forecast 853-8264 or case.jonathan(äensco.com for more responsibility lie within Brevard County on information on this work. Florida's east-central coast. Dr. Short generated the Thunderstorm Hail Index Probability Study bulletin using 14 years of (Dr. Short and Mr. Wheeler) RAOBs from XMR and the operational computer code WVTHUV1.f. The RAOB with a time closest The 45 WS has an operational requirement to to 1000 UTC between 0800-1200 UTC was issue weather advisories for hail of any size and selected for analysis for each day of the potential for severe weather elements, a subset of which is 2142-day database. The climatology of forecast hail with a diameter ^ 0.75 in. These advisories hail sizes was based on 697 days with a are issued for KSC, CCAFS, Patrick Air Force thunderstorm probability of "yes" and is shown in Base, and the Melbourne International Airport to Table 1. Also shown is the number of large hail protect personnel and material assets. The size reports on 42 hail days in Brevard County forecasters must also provide the probability of during the same reporting period. The large hail hail at any of these locations for the day at the size reports were obtained from the National 0700L weather briefing. The 45 WS tasked the Climatic Data Center's Storm Reports website. AMU to evaluate the current operational tools Table 1. Climatology of hail sizes forecast from the XMR morning RAOB and observed in Brevard County for May-September for the years 1989-2002. Description Approximate Size (inches) # Forecast # Obseried >Golfball ^2 1 1 Ping Pong/Golfball 1.5 - 1.75 1 7 Quarter to Half Dollar 1- 1.25 2 13 Dime/Nickel 0.75 - 0.88 0 21 Marble 0.5 0 N/A Pea 0.25 575 N/A <Pea <0.25 118 N/A Corrected Climatology of Forecast Hail Sizes height of the wet-bulb zero temperature was from the 45 WS "Thunderstorm Probability actually the height of the zero degree isotherm. Study" Mr. Kulow of CSR, who runs the program, concurred with the error analysis and forwarded Table 1 reveals a significant discrepancy remedial recommendations to CSR. between the number of forecast days with large hail (4 ^ 0.75 in) and the number of observed Dr. Short made changes to the AMU version days(42 ^ 0.75 in). The occurrence of large hail is of WVTHUV1.f to correct for the CCL and the under-forecast by a factor of 10. Because the height of the wet-bulb zero temperature and forecast hail size depends on a computerized regenerated the climatology of forecast hail sizes. analysis of the temperature and humidity profiles Table 2 shows the resulting corrected climatology from the morning RAOB, Dr. Short examined the of forecast hail sizes. WVTHUV1.f computer code in an effort to Contact Dr. Short at 321-853-8105 or determine the source of the discrepancies short.david.ensco.com, or Mr. Wheeler at 321- between the forecast and observed hail sizes. He 853-8205 or wheeler.markensco.com for more found the following two coding errors: 1) The information on this work. temperature of the convective condensation level (CCL) was much too high; and 2) the computed Table 2. Climatology of corrected hail sizes forecast from the XMR morning RAOB and observed in Brevard County for May-September for the years 1989-2002. Description Approximate Size (inches) # Forecast # Obseried j_ > Golfball ^ 2 29 1 Ping Pong/Golfball 1.5 - 1.75 7 7 Quarter to Half Dollar 1 - 1.25 28 13 Dime/Nickel 0.75 - 0.88 24 21 Marble 0.5 38 N/A Pea 0.25 454 N/A <Pea <0.25 117 N/A