Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Observing Systems The weather sensitivities of the space program demand mea- surements of parameters quite different from those made for use in providing the public with weather forecasts. The types and sizes of precipitation particles in clouds and the potential for triggered lightning are just two examples. Because of the special requirements of the space program, certain deficiencies exist in the observational program at KSC and other sites that can be remedied by a combi- nation of upgrading existing systems, acquiring and deploying new equipment now available, and conducting applied research to develop needed equipment not yet available anywhere in the world. These activities range from adding displays and calibrating instruments, which could be accomplished in a few days, to applied research that could take a few years. Improvements should be planned and coor- dinated by the Weather Support Office (WSO). Most of the critical weather elements discussed in Chapter 1 cannot currently be observed with the high degree of accuracy re- quired in an endeavor as weather-sensitive as the space program, where small errors can produce catastrophic results. Although most public-service forecasters would be pleased to be correct 90 percent of the time in yes no forecasts of precipitation, an accuracy that low for any of the weather elements critical for space flight could be devastating. The inescapable conclusion is that accuracies of about 99 percent or greater are needed when critical failures would result. 27
28 This requirement almost certainly dictates that Sections concerning weather-sensitive operations (1) will always be made as late as possi- ble, (2) will be based largely upon observations at decision time, and (3) should err in favor of postponing the weather-sensitive activity if critical weather is even a slight possibility. Thus, aside from planning efforts that require forecasts for days or longer, the forms of weather information most important for space operations are diagnoses of ex- isting conditions and very-short-term weather forecasts for periods of several hours or less. As long as launches are infrequent and delays are tolerable, there is likely to be little pressure on the system. However, as launches become more frequent, weather-related delays will be less tolerable, and therefore improved capabilities for detection and forecasting of adverse weather are needed. How unfailingly can state-of-the-science instruments adequately detect critical weather elements? How well can state-of-the-science methods be used to forecast critical weather elements for 2-hour intervals? This chapter gives an overview of (1) some of the existing mea- surement systems used at KSC (and, to a limited extent, at other sites), (2) other systems available for deployment, and (3) remaining needs for development of instrumentation to observe a few important meteorological parameters. UPP1:R-AIR SOUNDINGS High-resolution vertical profiles of wind speed and direction are needed to assess wind loads on the launch vehicle during launch and landing. The jimsphere balloon, tracked by radar, provides the greatest vertical resolution in measuring winds aloft. Data are normally obtained at 100 foot (atom) intervals. Jimspheres provide the data used in assessing the wine! loads prior to launches at KSC and Vandenberg. Near the jet stream there can be large wind variations in less than 2 hours that could male prelaunch balloon-based soundings unrepresentative of launch conditions. Balloon-based wind profiles require about an hour to measure winds to 55,000-foot (17 km) altitudes, so it is impossible to obtain soundings at less than 1-hour intervals unless multiple tracking devices are available and several balloons are airborne at the same time. Doppler wind profilers, which have been under development for a
29 decade, are in operation in a number of places worldwide. Although their vertical resolution is somewhat poorer than that of the jim- sphere system, wind profilers can provide data at intervals as short as 30 seconds, if desired. There are plans to install a Doppler wind profiler at KSC before the end of 1988. The wind profilers should be installed at and surrounding KSC in order to monitor important changes in the wind. Wind and wind shear data, as well as spectrum width of the profiler winds (which is related in part to turbulence within the beam), should be collected. Once a suitable profiler data base ~ attained, the method of as- sessing launch wind load hazards to the shuttle should be examined. It should be determined if a network of wind profilers at and sur- rounding KSC could be used to obtain very-short-term forecasts of wind profiles at launch tune through advection of wind field patterns across the network. A numerical mode! might be helpful in making these forecasts. The type of wind data really needed during a launch is a profile along the launch trajectory. Neither balloons, which drift with the wind, nor profilers can provide this type of sounding. Aircraft are better suited to provide this type of information, but the present prelaunch aircraft are not instrumented to make accurate wind mea- surements. The program of prelaunch reconnaissance flights using T-38 and Shuttle Training Aircraft should be upgraded either by adding instrumentation to these aircraft or by using other available instrumented aircraft. Quantitative measurements should be made, over and upwind of KSC, of cloud electric fields, the types and sizes of precipitation, electric fields and Maxwell currents, winds, wind shears, and turbulence. A computerized data collection system should be used to facilitate the real-time collection and archiving of these data, and also to transmit the data to KSC forecasters for timely use. MSFC should explore the possibility of using these data as part of the DISC loads assessment program. Thermodynamic soundings (temperature and relative humidity) are needed to obtain atmospheric density profiles during launches. These are obtained by balloon-based instruments, particularly the ground meteorological detector (GMD)-tracked radiosondes, and by rocketsondes. These systems should be assessed against the state of-the-science technology, such as Loran-based balloon tracking sys- tems. The latter have proven far superior to GMD systems for obtaining accurate wind speed profiles during field research experi- ments, especially during situations of strong winds aloft and in terms
30 of vertical resolution. Furthermore, the National Center for Atmo- spheric Research (NCAR) Cros~cha~n Loran Atmospheric Sounding System (CLASS) has been designed and demonstrated to operate nearly automatically, and would potentially provide better data with less manpower and cost than the present GMD system. Remote soundings of temperature and humidity, obtained via satellite-based radiometric profiling, currently have vertical resolu- tion that ~ too coarse for use in the space program. WSO should monitor the progress of research on these systems and be prepared to put them into use in the space program, should their resolution improve. To obtain better htformation about spatial and temporal variation of the wind near ESC, NASA shoed establish a network of Doppler wind profilers and a program for en- hanced aircraft observations mmg avaBable NASA and U.S. Air Force aircraft. BOUNDARY LAYER AND SURFACE WEATHER Near-surface winds are important for landings, launches, and ground operations, and can be measured accurately and at very fre- quent intervals (1 minute or less) using automated weather stations. A system of this type, called WINDS (Weather Information Network Display System), is used at both KSC and Vandenberg, with sen- sors on towers from 54 to 500 feet at KSC and 12 to 300 feet at Vandenberg. Winds from these networks are available at A, 15-, or 3~rntnute intervals. Shorter intervals may be desired in the critical minutes before launch, when passage of a gust front (outflow from a distant thunderstorm) or the sea breeze front (moving in from sea) could cause dramatic changes of wind direction and speed. The existing automated surface mesonetwork (28 stations) is a critical element in the observational program at KSC. It should be expanded to the west to cover the western portions of the KSC activity domain (and procurement of 20 additional stations is in progress), and to the east to include measurements over water, via buoys or platforms for routine operations and/or via ships during launch situations. The instrumentation should be expanded to in- clude visual range transm~ssometers at the launch pads and the Shut- tle landing field airstrip. The individual sites should be adjusted, if
31 necessary, to ensure that the observations are taken at uniform al- titudes, with proper exposure and sheltering, and with uniform and well-ma~nta~ned instrumentation. A Doppler sodar (sonic detection and ranging) can be used to monitor the low-level (up to about 1 km) wind profile at 5-minute interval except during precipitation. This instrument has better ver- tical resolution than the wind profiler, 80 Doppler sodars would be of value in augmenting the tower wind network. Such data would prove invaluable for dispersion forecasting and in providing information regarding other surface operations. A Doppler Acoustic Sounding System (DASS) is currently operated at Vandenberg. The horizontal distribution of low-level winds provides important information for weather forecasting. Small-scale fronts and wind shift lines can escape detection if stations in a mesonetwork are more than several kilometers apart. Scanning Doppler weather raciars and Doppler lidars can supply the type of spatial coverage needed to locate such wind shift lines. A NEXRAD Doppler racier ~ expected to be installed at Melbourne, Florida, about 25 miles south of KSC, in 1990. Because a single Doppler radar can detect motion only along a radial, a network of at least two Doppler radars should be deployed at KSC ~ order to resolve tote] horizontal velocities. Unfortunately, the NEXRAD radar to be deployed at Melbourne within the next several years will not scan in a manner conducive to multiple Doppler radar studies in consort with another radar. NASA should acquire at least two dedicated Doppler radars, which would enable calculation of detailed patterns of winds in clouds and in the boundary layer. To make the wind calculations in real time would require the develop ment of new dual-Doppler data processing and display software. In addition to horizontal mappings of velocity, cros~sections along the space vehicle flight path could also be constructed. 1~ obtain enhanced information about low-le~re} wmds and other weather elements, NASA should emend the areal cov- erage of the surface mesonetwork and Include data platforms over the ocean. At least two dedicated Doppler radars shown be installed in locations that optimize coverage over KSC to improve forecasts 1lemg higher resolution boundary layer data and to better relate the wmd fields and refiectivity withm clouds to the microphysical and electrical develop meet. NASA Should consider deploying Doppler sodars for mowtoring the boundary layer.
32 PRECIPITATION Showery precipitation often falls over areas of only a few square kilometers, and rain gauge networks are rarely dense enough to resolve this detail. Conventional (non-Doppler, incoherent) weather radar can be used to obtain high-resolution mappings of areas with precipitation. Forecasters use the horizontal and vertical shapes of the radar "echoes and the intensity of the echoes to identify convective and stratiform precipitation. Satellite imagery can also be used to help identify convective clouds. However, neither radar nor satellites can unambiguously distinguish thunderstorms from other types of convective precipitation. State-of-the-science weather radars provide digital data that can be processed by computerized software packages to derive additional useful products such as vertically integrated liquid water contents, cross sections of reflectivity at any desired angle, and animated im- agery. The 30~year-old FPS-77 radar at Vandenberg is not digitized and provides the forecaster only with snapshot views at fixed az- imuth or elevation angle. A radar should be deployed at Edwards AFB, and digital radars should be considered for both Vandenberg and Edwards. The thermal tiles on the Space Shuttle are eroded by precipita- tion drops. However, there is a need for more detailed information relating drop size and concentration to the extent of the tile damage. Unfortunately, drop sizes cannot be measured using conventional radar. Surface-based disdrometers are typically used to measure raindrops reaching the ground, and an aircraft-mounted Knollenberg probe can be used to sample sizes of precipitation aloft. These types of instrumentation are not currently used in space operations, but should be. A possible tool of the future is the multiparameter radar, which transmits at two wavelengths and with two polarities. Multiparam- eter radars can distinguish between snowflakes, raindrops, and hail, and between large drops and small drops. However, certain ambigui- ties exist, such as melting snowflakes. Additional research should be done to enable this tool to be utilized operationally. To obtain data on cloud and precipitation types and sizes, air- borne tIrop-size measuring instrumentation should be flown prior to Space Shuttle launches, and a mo~tiparameter radar should be acquired.
33 LIGHTNING During the summer at KSC there is an average of about six lightning strikes to ground per square mile each month. Until the last decade, it was extremely difficult to detect and locate lightning strikes on a real-tune basis. Cloud-to-ground lightning strikes can now be successfully detected by using either magnetic direction- finding (Lightning Location and Protection (ALP)), omnidirectional broad-band time-of-arrival (TOA) antennae (Lightning Position and Tracking System (LPATS)), or by careful interpretation of electric field mill network data. (Other methods also exist, such as lightning- detection ra`lar and lightning interferometers.) Lightning strikes are typically located with position accuracies of 2 km or better by triangulation. An LLP system is in operation at KSC; it should be improved by periodically checking the site correction factors and the antenna alignments. Two larger lightning detection networks cover the KSC area: a network of LLP direction finders operated by the State University of New York at Albany and the Florida LPATS network of broad- band TOA receivers. Displays of these data should be added to the KSC weather office. Data from the SUNY Albany system showed the movement of an area of considerable cloud-to-ground lightning activity toward KSC from the west prior to the AtIa+Centaur 67 accident, as shown in Figure 4. Had these data been available in the KSC weather office, it is likely that the launch would have been postponed, averting the accident. At KSC, at present, in-cloud and cloud-to-cloud lightning dim charges are difficult to detect. These occurrences can be inferred from data provided by the Launch Pad Lightning Warning System (LPEWS), a Gestation network of field mills that is clesigned to de- tect electrified clouds. Because the LPEWS is the only network of its kind in the world, few meteorologists have been exposed to these data for use in real-time weather analysis and forecasting. Persons who would typically provide forecaster training are not usually well versed in this tool, and those familiar with field malt network inter- pretation are usually more adept at using it in a research rather than an operational environment. The LPL.WS is currently being upgraded. The sensors should be improved, and the sites should be carefully evaluated to identify any local obstructions or sources of contamination, and obstructions should be removed or sites relocated, if necessary. The network should be expanded to the west and to the east, including over-water
34 ..- - I_ \~j.~;~. U .- ~.~' ~ -- eF ~ ~, , ~ _ :~ _ ~ ___ ~_~- r ~`LAUNCH , SITE _ _ = e _ R ~_ TIMES UTC . ·1823-1923 ·1923-2023 ·20 23-21 23 FIGURE 4 State University of New York (SUNY) at Albany display of LLP- detected cloud-to-ground lightning prior to the Atlas-Centaur 67 launch. In the 3 hours prior to launch, lightning activity progressed steadily across Florida toward KSC. (Courtesy of R. Orville, State University of New York at Albany.) sites. The equipment should be carefully calibrated and certified for operational use, and the observations included in the list of weather criteria for launch (and landing). In-cloud and cloud-to-cloud lightning can also be detected by using networks of (1) HF or VHF time-of-arrival receivers or (2) HF or VHF lightning interferometers. A system (LDAR) of the former type was previously operated at KSC but abandoned. A new system of this type should be built. The National Aeronautics and Space Administration should make improvements to the existing LIP and [PEWS systems and obtain displays of other lightning detection networks in the area, ~ order to improve detection of lightning and elec- tric fields. A new system should be blight to detect lightning in and between clouds aloft.
35 DCLOU ELECT lIC FI1:IDS Clouds, such as thunderstorm anvils, stratiform thunderstorm anvils, stratiform clouds, and shallow convective clouds, often do not produce lightning but do contain high electric fields. The threat of triggered lightning from these clouds may be the most difficult weather hazard to detect and forecast. Surface electric fields do not always reveal electric fields aloft or charge centers in the upper por- tions of clouds, because of the presence of intervening (or screening) charged layers. Airborne electric field mild systems, such as those formerly used on the NASA F10~B research aircraft, should be used to accurately characterize the electrical environment aloft. Much of the data collection and research on the subject of trig- gered lightning has been sponsored by KSC, so that the center's triggered lightning research is state-of-the-science within the atmo- spheric electricity community. Additional efforts are needed to add companion meteorological data (such as radar data, surface mesonet- work and tower data, satellite data, and sounding data) to the trig- gered lightning data base for possible forecasting applications and to provide training to operational forecasters concerning the use of field malt network data. Airborne measurements using field mats repro sent an import ant contribution to better defining the potential for induced lightning. The new launch criteria, designed to avoid any possibility of triggered lightning, may have become overly conservative with re- gard to cloud electric fields. To addre" this issue, one or more instrumented aircraft should be flown on frequent occasions in order to develop a cI~matolog~cal data base reg=ding electric fields and Maxwell currents in ~dead" or detached anvils and anvils from dim t ant thunderstorms. Data should also be collected in other types of cloud near the freezing level. Triggered lightning studies shown be continued, with addi- tional efforts to collect companion meteorological data sets. A'rbo~e electric field measurements shown be collected to enhance studies of the threat of triggered lightumg. OTHER WEATlIER ELEMENTS Dangerous icing conditions will result if a vehicle encounters supercooled (i.e., liquid at subfreezing temperatures) cloud and pre- cipitation drops. Owing to the poor spatial coverage of rawinsonde
36 data and since conventional weather radar cannot distinguish be- tween precipitation sizes or types, regions conducive to aircraft icing are very difficult to detect. Pilot reports are the main source of infor- mation. Multipararneter radar, combined with temperature profiles, might prove useful for detecting and avoiding freezing rain. Cloud radars (wavelength of approximately 3 mm), which detect cloud- sized particles, may prove useful in supercooled cloud detection, if used with sounding data. Clear-a~r turbulence, which arises within layers of large vertical wind shear, ~ very hard to detect. It is most commonly detected and reported by pilots. Some information regarding shears and hence the possibility of turbulence can be derived from the spectrum width of Doppler radar data, both from scanning Doppler weather radar and from Doppler radar wind profilers. Much work remains to be done, however, in calibrating the spectrum width values against the incidence of turbulence. Satellite imagery can also often be used to alert forecasters to areas where turbulence is likely. Trained weather observers can also provide valuable data to the weather forecaster. An observer has the unique ability to assimilate audible and visual data in a manner that is better than most instru- ments. To obtain quality information, the observers must be trained to identify the specific conditions that may be conducive to weather hazards such as triggered lightning. At KSC, the weather office has no windows, and forecasters cannot see outside without climbing to the roof. It would be desirable to move the forecasting operations to a room with a window or to make a window in the room currently used, so that observers could more easily monitor rapidly changing atmospheric conditions. Trained and reliable observers and adequate facilities are needed at all sites overseas and in the United States. The pane! does not feel comfortable with past arrangements for obtaining weather observa- tions at overseas landing sites. The National Aeronautics and Space Administration should ascertain that launch and landing sites are provided with skilled observers and necessary measurement systems. NASA should monitor the achievements in observation technology and deploy usefid new instrumentation expediently.