• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Aviation Applications for Satellite-Based Observations of Cloud Properties, Convection Initiation, In-Flight Icing, Turbulence, and Volcanic Ash

Posted on: Monday, 12 November 2007, 09:00 CST

By Mecikalski, John R Feltz, Wayne F; Murray, John J; Johnson, David B; Bedka, Kristopher M; Bedka, Sarah T; Wimmers, Anthony J; Pavolonis, Michael; Berendes, Todd A; Haggerty, Julie; Minnis, Pat; Bernstein, Ben; Williams, Earle

Aviation safety is being advanced as meteorological satellites are exploited for critical information. Turbulence, in-flight icing, convective storms, and volcanic ash hazards are being mitigated as satellite parameters are implemented within national aviation forecasting systems. Weather hazards can reduce the safety and constrain the efficiency of our national air navigation system. Not only are most of the air traffic delays caused by weather, but weather is also a factor in many fatal accidents. The Federal Aviation Administration's (FAA's) National Aviation Safety Data Analysis Center (NASDAC), in a study of the National Transportation Safety Board (NTSB) accident database (a summary of the study's conclusions is available either at www.nasdac.faa. gov/ aviation_studies/weather_study/summary.html or from the Office of Systems Safety, Federal Aviation Administration, 800 Independence Avenue, SW, Washington, DC 20591), discovered that in the 10 yr prior to 2003, there were 19,562 aircraft accidents involving 19,823 aircraft. Weather was a contributing or causal factor in 4,159 (21.3%) of these accidents, with 41.2% of all weather-related accidents having no record of the pilot receiving a weather briefing. Small general aviation aircraft experience the vast majority of accidents and fatalities, with wind gusts, low ceilings, and limited visibility being the primary weather causes. Other critical factors affecting both small and large planes include turbulence, wind shear, and in-flight icing. Turbulence alone has been estimated to cost the commercial airline industry about $750 million yr', with "low end" general aviation aircraft having the largest negative safety impact for these hazards (Triplett 2005). Other significant hazards to aviation that cause significant travel delays and/or are causes for significant concern include in-flight icing and suspended volcanic ash. Over this 10-yr period, the annual number of weather-related accidents has been declining, but has remained roughly constant as a percentage of the total number of accidents.

From the earliest days of aviation it was clear that high- quality weather observations and accurate weather forecasts were critical to the safe operation of aircraft. In the early twentieth century the development of a national meteorological surface- observing system paralleled the expansion of the aviation industry. After World War II, weather radars gradually came into widespread use for detecting the location, height, and movement of major storm systems for general public safety and support of aviation. In the late 1970s, the Center Weather Service Units of the Air Route Traffic Control Centers were formed, following an airline crash in north Georgia. In the 1980s the National Weather Service (NWS) began a nationwide modernization program, in partnership with the FAA, to upgrade weather-observing systems, including next-generation weather radars [Weather Surveillance Radars-1988 Doppler (WSR-88D)] and automated surface-observing systems (National Research Council 1980, 1991, 1994).

The movement toward specialized meteorological observing systems to support aviation was accelerated by T. Fujita's identification of small-scale wind patterns from convective storms that caused a number of major airline crashes (Fujita and Byers 1977; Fujita and Caracena 1977; Fujita 1985, 1986). The need for operational detection of these wind patterns led the FAA to install specially designed Terminal Doppler Weather Radars (TDWR) or Low-Level Windshear Alert Systems (LLWAS) at many of the major airports in the United States. The parallel need for high-resolution, rapidly updated aviation weather forecasts led to the development of special numerical weather prediction (NWP) models, such as the National Center for Environmental Prediction's (NCEP's) Rapid Update Cycle (RUC) model.

During this same time period, satellite-observing systems were also undergoing a rapid expansion in their capabilities and availability and represented an underutilized resource for enhancing aviation weather products. Recognizing that advanced satellite capabilities could contribute to aviation safety and efficiency, the National Aeronautics and Space Administration (NASA) developed an Advanced Satellite Aviation-Weather Products (ASAP) initiative in 2002 to help transfer some of the new satellite-observing systems and products into operational use through a collaborative effort with the FAA's existing Aviation Weather Research Program (AWRP).

The AWRP focuses on developing and enhancing aviation weather products that respond to the needs of users and FAA priorities. All new weather products developed by the AWRP's Product Development Teams (PDTs) go through extensive testing and evaluation, culminating in a formal review by a joint FAA and NWS Aviation Weather Technology Transfer Board (AWTT) before they are declared operational and made available to the public. The AWTT review process puts an equal emphasis on high accuracy and low false-alarm rates.

The NASA-funded ASAP initiative is structured to work with the AWRP's existing PDTs to help fill critical gaps or to extend current products, while making sure that the FAA is aware of the new capabilities that are becoming available and may be ready for operational use (see Mecikalski et al. 2002). In general, the ASAP contributions to the PDTs will not result in stand-alone "satellite" products, but rather contributions toward integrated products that also incorporate multiple observing systems and NWP model results.

Our paper highlights this highly complementary collaboration, beginning with an overview of the initiative in "ASAP overview," and continuing with descriptions of the individual products in later sections.

ASAP OVERVIEW. ASAP is administered by the NASA Langley Research Center (LaRC) in close collaboration with the FAA's management of its AWRP efforts. The ASAP program is sponsored by the NASA Aviation Safety and security Program (AvSSP) and the NASA Applied Sciences Program's Aviation Applications Program, whose goals to achieve broader societal impacts are also closely coordinated with the FAA (Murray 2002). Like the AWRP team members, the ASAP team represents a collaboration between university and government researchers, with a specific goal of bringing together groups that are experienced in satellite remote sensing, but who have not had a history of working on aviation-related problems, with researchers having long histories of studying aviation weather.

Current participants in the ASAP program include scientists and engineers at NASA's LaRC, the University of Alabama in Huntsville (UAH), the National Oceanic and Atmospheric Administration (NOAA) Cooperative Institute for Meteorological Satellite Studies at the University of Wisconsin (UW-CIMSS), the Massachusetts Institute of Technology Lincoln Laboratory (MIT-LL), and the National Center for Atmospheric Research (NCAR). [See ASAP Web sites at cimss.ssec.wisc.edu/asap/ (UW-CIMSS) and nsstc. uah.edu/johnm/ ci_studies.html (UAH); the NASA ASAP page is http:// science.larc.nasa.gov/asap/index. html.]

The ASAP effort is envisioned as two complementary phases. The first phase emphasizes enhancing the FAA's AWRP use of current satellite data and processing techniques, while the second phase will seek to identify aviation applications for the next generation of high spectral and spatial resolution satellite instruments and give aviation requirements a higher priority in future satellite planning. The appendix outlines several of the anticipated advanced satellite instruments that will have the capability for improving aviation weather diagnosis and forecasting, especially as aviation requirements gain a higher priority in future satellite planning. New satellite names and satellite-borne sensing systems that are not included in the American Meteorological Society's (AMS's) Glossary of Meteorology (Glickman 2000) are provided in the appendix.

In conjunction with the FAA's Product Development Teams, ASAP is working to both identify areas where satellite data can fill gaps in weather products already under development and propose new satellitebased "interest fields" that can be incorporated into integrated FAA weather products. ASAP includes efforts to improve cloud identification and properties, monitor the initial stages of developing convection, improve assessments of in-flight icing potential, identify flight areas that are likely to be turbulent, and develop automatic algorithms to detect volcanic eruptions and identify hazardous areas of volcanic ash.

BASIC CLOUD PROPERTIES. Satellite products have become widely available and are an essential component of most modern meteorological analyses. For the most part, however, satellite datasets are only used in the form of simple cloud masks (cloud or no cloud), cloud-top heights, or other basic properties. Clouds can also be tracked for making wind estimates (Velden et al. 1997, 1998). While these seemingly simple parameters are widely used, improved satellite-observing systems and new analysis techniques (e.g., Velden et al. 2005) can be combined to produce enhanced basic products (cloud masks, cloud classification analyses, and cloud-top heights) that can be transferred directly, and quickly, to existing operational analysis systems. While accurate observations of cloud- top height are critical for many aviation applications, the specific products that are needed can vary from a simple statute height above ground level (or sea level), to a separate cloud-top pressure product, which is appropriate for en route aircraft flying above 18,000 ft (5625 m). These aircraft, which set their altimeters to a standard reference, display their flight levels in thousands of feet while they are actually flying along constant pressure surfaces.

For some time, satellite-based infrared (IR) sounders have been able to generate operational cloud-top pressure fields through a multispectral CO2-slicing technique (Wylie and Menzel 1989). These fields, however, have not been used to any great extent for aviation applications because of the relatively large sounder footprint along with the product's limited area coverage and slow update rates. With the launch of the Geostationary Operational Environmental Satellite- 12 (GOES-12) with a new 13.3-[mu]m CO2-imaging channel, it has become possible to generate cloud-top pressure fields using the higher-resolution, rapidly updating GOES imager. In this case, the CO2-slicing technique has been adapted for use with the imager in a new CO2 ratio technique (Schreiner and Schmidt 2001; Bedka et al. 2007).

One of the advantages of the new CO2 ratio technique is to produce a better estimate of the height of thin, semitransparent cirrus and other high clouds that are difficult to assign to a specific height because of contamination of the cloud information by the underlying surface temperature field. Unfortunately, the CO2 channel is limited to mid- and upper-level clouds. A full cloud-top pressure product is therefore developed through the combination of height estimates from the CO2 ratio technique for the upper clouds and from the older "IR window" technique for lower clouds (Schreiner et al. 1993). The IR window technique uses the 11-[mu]m thermal IR temperature to estimate the temperature at cloud top, combined with an atmospheric sounding to convert the temperature value to a corresponding pressure and height. The AWRP oceanic weather PDT's current en route cloud-top height product (CTOP) is based on the IR window technique (Donovan et al. 2007).

Figure 1 shows an example of this GOES imagerbased cloud-top pressure product. Although optically thin clouds do not in themselves pose a major hazard to aviation, it is still important to report accurate cloud-top heights. Using a combined IR window and CO2 ratio technique also provides an added benefit over a single- channel technique. The CO2 ratio technique allows for an assessment of the emissivity of a cloud, while the IR window technique automatically assumes that this emissivity is 1 (as it would be for an optically thick cloud). This allows clouds to be classified as optically thick or thin with a higher degree of certainty, and prevents some cirrus clouds from being mistakenly classified as lower, optically thick clouds, which could be considered an aviation hazard.

Figures 2 and 3 show a detailed comparison between the GOES-12 imager and sounder cloud-top heights (converted to statute height in this example), coupled with simultaneous observations by the NASA Cloud Physics Lidar (CPL), flying on the NASA Earth Resources-2 (ER- 2) aircraft. These measurements were conducted in the winter of 2003 as part of The Observing System Research and Predictability Experiment (THORPEX) Regional Campaign in the Atlantic, known as ATReC. Figure 2 shows the GOES-12 imager (top, red) and GOES-12 sounder (bottom, blue), as well as CPL (black) cloud-top heights from 5 December 2003. The imager heights show better agreement with CPL than the sounder heights because of the increased spatial resolution of the imager. The best agreement for both satellite instruments is for midlevel clouds, while both the sounder and imager underestimate the CPL cloudtop height both for semitransparent high clouds, and warm low-level clouds.

Figure 3 shows a histogram of the absolute differences between the CPL cloud-top height and the GOES imager (red) and sounder (blue) cloud-top height for 5 December 2003. This histogram indicates that the distribution of differences between both GOES products and the CPL is bell shaped, with the peak in the 0.5-1-km bin. The distribution of differences for both satellite instruments shows a longer tail to the right, indicating that the satellites underestimate the CPL cloud-top height more often than they overestimate it. For both instruments, a similar percentage (-5.5% and 4.9% for the sounder and imager, respectively) of the data points had absolute errors > 5 km (outside the histogram range). This result suggests that the cloud types for which height assignments are most difficult (i.e., optically thin, broken, or multilayered clouds) are consistent for both broadband instruments. The root-mean-square and bias statistics for each satellite (summarized on the figure) confirm the better performance of the imager-derived product than the corresponding sounder-derived product.

CONVECTIVE INITIATION. Thunderstorms account for most of the air traffic delays in the United States and cost the aviation industry many millions of dollars annually in lost time, fuel, and efficiency through delayed, canceled, and rerouted flights (Kaplan et al. 1999; Murray 2002). In response to this need, the FAA's convective weather PDT is working to develop reliable 0-8-h probability forecasts of convective weather and automated 0-2-h high-resolution deterministic forecasts over the continental United States (CONUS).

Historically, much of the convective weather PDT's work has centered on the use of weather radars to monitor convective storm development (Mueller et al. 2003). Radars, however, do not see the early stages of cumulus cloud development prior to precipitation development. In the late 1990s, NCAR began to incorporate satellite information into the 0-1-h convective storm nowcasting system (the Auto-Nowcaster). Satellite feature detection algorithms (e.g., Bankert 1994; Roberts et al. 1999; Tag et al. 2000, and Roberts and Rutledge 2003) were used to classify cloud types, identify early cumulus growth above surface convergence boundaries, and monitor the growth of these clouds by observing the changes in their IR cloud- top temperatures. The ASAP team is currently working with the FAA convective weather PDT to improve methodologies for satellite-based techniques for detecting, tracking, and monitoring the early development of small cumulus in order to improve thunderstorm initiation forecasts over the CONUS. Future efforts will explore the transfer to ASAP assets to oceanic and remote domains.

The first step in the ASAP satellite-based convective initiation (CI) product is a new convective cloud identification system developed at UAH and UW-CIMSS, the "convective cloud mask" ("CCM"). The ASAP CCM relies on an unsupervised statistical clustering algorithm to classify the cloud features in satellite imagery based on the methodology of Nair et al. (1998). The iterative clustering algorithm groups pixels into clusters based upon a statistical similarity measure (Levine and Shaheen 1981). The clustering algorithm is trained using GOES data over the continental United States. All of the GOES channels were used as features. In addition, the contrast texture measure of channel 1 [visible (VIS)] is used as a training feature for the clustering algorithm.

While the CCM can be used with both geostationary and polar- orbiting satellite systems, its application to CI requires monitoring the cloud imagery over time and is best suited for use with GOES imagery updated every 15-30 min. The CCM is optimized for identifying small convective clouds through the use of both 1-km- resolution VIS and 4-km-resolution IR imagery, with the CCM being produced at 1-km resolution. This means that the CCM is only available during daylight hours. In spite of its high resolution, the CCM's emphasis on nonprecipitating convective clouds simplifies the processing and allows it to be efficiently applied over large geographical areas.

Figure 4 shows the clustering results for a GOES image over the central United States. Large-scale convective complexes are visible, along with smaller cumulus and stratiform clouds. The classifier is able to detect the smaller convective clouds. Additionally, the isolated overshooting cumulus tops can be seen in red. These overshooting tops are identified by the use of a pseudotextural channel, created from GOES VIS and IR imagery (see Nair et al. 1998); the texture allows for the delineation of convective cloud features among a shield of status-like clouds, in this case, the stratus clouds are cirrus anvils.

Convective clouds identified by the CCM are then tracked using a modified version of the Velden et al. (1997, 1998, 2005) atmospheric motion vector (AMV) algorithm. The modified version is optimized for mesoscale studies by reducing its dependence on first-guess motion estimates from NWP models. This approach results in a higher density of cloud motion vectors that includes both synoptic and mesoscale flow components, as discussed in Bedka and Mecikalski (2005).

Motion vectors are computed by identifying cloud "targets" across an image sequence through the use of advanced pattern-matching techniques and cross-correlation statistics. With the relaxed constraints to synoptic-scale forcing, the number of AMVs calculated can increase significantly (as much as a 20-fold increase in motion vectors that pass quality control filtering), giving a better depiction of the mesoscale flow patterns that can influence CI. Figure 5 shows the increased coverage of the height-stratified satellite-derived motion vectors using this analysis technique. It is relatively easy to identify and track large cloud features by satellite. These cloud features, however, are usually well tracked by weather radars. The critical contribution of satellite AMVs to predicting early thunderstorm development is to be able to track the growth and evolution of the smaller-scale cumulus and lines before they develop precipitation echoes. Tracking these small cloud features, in turn, requires very accurate knowledge of the low- level wind field. By combining the CCM and the corresponding AMVs, it is possible to monitor the changing cloud properties for individual, moving small cumulus with appropriate indicators for inaccurate advective motions applied to cloud motions (see Mecikalski and Bedka 2006). These tendencies can then be monitored for signals of cloud growth. Cloud-top cooling rates, for example, are a good general indication of cloud development. Roberts and Rutledge (2003) showed that the occurrence of subfreezing 10.7- [mu]m brightness temperatures (T^sub B^), in the range from 0[degrees] to -2O0C, coupled with cooling rates of 8[degrees]C in 15 min, provided up to 30 min of advance notice of the development of cumulonimbus with radar reflectivities of 35 dBZ or greater. Figure 6 illustrates the calculated cooling rates for areas of developing convection in the Midwest.

Mecikalski and Bedka (2006) have extended these tendency techniques to a variety of multispectral band differences from GOES imager data to produce eight IR predictor fields that can be used to characterize conditions consistent with CI for moving cumuli. The fields are 1) 10.7-[mu]m cloud-top T^sub B^, cloud-top T^sub B^ cooling rates at 2) 15 and 3) 30 min, 4) cloud-top T^sub B^ temperatures dropping below 0[degrees]C, the 5-6) 6.5-10.7-^m channel difference as discussed by Ackerman (1996) and its time rate of change, and the 7-8) 13.3-10.7-[mu]m channel difference and its time rate of change. After selecting suitable thresholds for these growth trend indicators, Mecikalski and Bedka generate a satellite- based, summary CI product defined by small clouds that meet at least seven of the eight growth criteria. The results of this type of analysis are shown in Fig. 7, where the red pixels represent small convective clouds that have been identified as being likely to develop >/=35-dBZ radar echoes within the next hour. The FAA's convective weather PDT is evaluating this satellite-based CI product, along with its individual trending components, for inclusion in their short-term (0-6 h) forecast systems (Cai et al. 2006). The NOAA Satellite Analysis Branch is performing a similar evaluation as of summer 2007 (J. Hanna 2006, personal communications). Figure 8 shows an example of the ASAP cloud-top cooling field as used within this forecast system.

Accuracies of ~55%-65% (in terms of "probability of detection") have been recently determined for the Mecikalski and Bedka (2006) method, and these will be reported in an upcoming paper. It was also determined that via principal component analysis, all eight CI indicators have unique value within the nowcasting procedure.

IN-FLIGHT ICING. Aircraft penetrations of supercooled clouds can result in a rapid buildup of ice on the aircraft wings and control surfaces. In extreme conditions this ice buildup has been directly responsible for airplane crashes. On 31 October 1994, for example, an ATR-72 commuter airline crashed in Roselawn, Indiana, killing all 68 persons on board, following a buildup of ice on the wings and airframe. While the microphysical conditions that result in inflight icing hazards are generally well understood, it has proved challenging to develop reliable techniques for operationally diagnosing icing conditions or forecasting their occurrence. For many years the FAA has sponsored the development of advanced techniques to identify icing hazards and alert pilots to the danger thereof. These techniques rely on a variety of data sources, including radar, satellite and surface observations, and pilot reports, as well as predictions from highresolution, rapidly updated NWP models. At present, the most sophisticated operational icing product is the Current Icing Potential (CIP; Bernstein et al. 2005), developed by the FAA's in-flight icing PDT, which is run operationally by the National Weather Service. The current operational version of CIP makes relatively limited use of satellite data, using multiple channels to develop a 3D cloud mask. The ASAP program intends to enhance CIP through the use of advanced satellite products.

Satellite-based methods using combinations of VIS reflectance and IR emittance can be used to estimate cloud properties, such as hydrometeor size and liquid water path (LWP). Ellrod and Nelson (1996), Vivekanandan et al. (1996), and Minnis et al. (2001) have suggested that in addition to characterizing the location, size, and depth of cloud systems, satellite observations can be used to monitor the cloud-top phase (i.e., ice or liquid), water content, and droplet diameter, all of which are properties that are of interest in diagnosing possible icing conditions. Although satellite- based cloud retrievals using the VIS and IR spectral bands can only monitor conditions over a limited depth at the top of the cloud, these techniques have shown skill in monitoring hydrometeor size and phase information in that region. In some cases the cloud-top regions may even be of special interest since largest amounts of supercooled liquid water are often found within several hundred meters of the top of layered stratiform and convective clouds (Hobbs and Rangno 1985; Rauber and Tokay 1991). While cloud-top information from satellite retrievals can be valuable, stand-alone satellite products, they must be used with caution since the hydrometeors at cloud top are not always typical of those below. For general use, satellite techniques need to be integrated with other observational systems and model output, such as those employed by CIP.

The ASAP icing team is currently evaluating GOES-derived cloud products developed at NASA LaRC (Minnis et al. 2001) for integration into the FAA's operational CIP system. These products are generated with the Visible Infrared Solar-Infrared Split-Window Technique (VISST) during the daytime and with the Solar-Infrared Split-Window Technique (SIST) at night (Minnis et al. 1995). Each 4-km GOES pixel is first classified as either clear or cloudy using a sophisticated cloud identification scheme (Trepte et al. 1999). Multispectral VISST/SIST algorithms are then applied to each cloud pixel to determine cloud phase, optical depth, effective particle size, effective temperature, effective height, and ice or liquid water path.

In collaboration with the FAA's in-flight icing PDT, the ASAP icing team has been testing NASA LaRC cloud products as a nowcasting tool for guiding research aircraft into areas with the greatest potential for icing. The satellite-derived cloud products, in conjunction with other standard meteorological information, have already been used in support of the NASA Glenn Research Center icing research flight program (Wolff et al. 2005) and the 2005 Alliance Icing Research Study (Isaac et al. 2005). In these experiments, satellitederived phase, LWP, and effective radius (R^sub e^) were used to identify areas with possible icing conditions to guide research aircraft to collect in situ observations. The aircraft observations, in turn, can be used to evaluate the NASA LaRC products.

For example, on 19 January 2005 forecasters targeted clouds over central and western Ohio because of a combination of supercooled liquid phase cloud tops, relatively high LWP values of 800-1000 g m^sup -2^, and R^sub e^ values on the order of 10 [mu]m (see Fig. 9). Widespread cloudiness covered the region on this day. The limited cloud-free areas in the figure are colored black. Clouds with glaciated tops are colored gray. Upon reaching the area, the aircraft encountered liquid water contents measuring 0.4-0.6 g m^sup -3^. Moderate icing was indeed observed on the aircraft during this flight, as well as by several others in the area. These values were very close to those anticipated from the satellite-derived liquid water path estimates, and in situ observations of mean volume drop size diameters (~15 [mu]m) were in good agreement with R^sub e^ estimates (~10 [mu]m). During earlier portions of this flight made along the Ohio-West Virginia border, liquid water contents observations did not match the values expected from the LWP estimates nearly as well (0.2-0.3 g m^sup -3^ with LWP of 700-1000 g m^sup -2^). This discrepancy may be explained by the presence of large R^sub e^ values (>20 [mu]m) at cloud top. Since the VISST algorithm calculates LWP using the cloud-top R^sub e^ value, it may overestimate LWP if the cloud-top R^sub e^ value is larger than particle sizes in lower parts of the cloud (Khaiyer et al. 2003). Note that the aircraft did encounter larger drop sizes during this portion of the flight.

After three seasons of using the NASA LaRC products, field program forecasters have found them to be quite useful and have developed their own "rules of thumb" for when specific products are of value. For example, large areas with uniform values of phase, LWP, or R^sub e^ are usually more reliable indicators of icing conditions than areas with high spatial variability. Gradients in LWP that are embedded within widespread stratiform clouds are usually found to be accurate at least in a qualitative sense, while sharp transitions are sometimes problematic. Cloud-top phase estimates are typically quite good, and higher R^sub e^ estimates correlate well with in situ observations of larger drops (Haggerty et al. 2005). The ASAP icing group is currently conducting more comprehensive validation studies using research aircraft data and pilot reports, as well as comparisons with the current generation of CIP output. The forecasters' observations and insights, in combination with these more comprehensive validation studies, are providing direction to the creation of specific rules and procedures for integrating the advanced satellite-derived measurement fields into a new, highresolution CIP product that will focus on expected icing severity.1 Advanced sensors on future satellites, including the National Polar-Orbiting Environmental Satellite System (NPOESS) and GOES-R, will enhance the capabilities for detecting large droplets, diagnosing icing conditions in multilayered clouds and at night, and estimating LWP over snow-covered surfaces. For example, the 1.6- and 2.2-[mu]m channels on the GOES-R Advanced Baseline Imager (ABI) can be used on conjunction with the 3.9-[mu]m channel to estimate the cloud vertical profile of droplet sizes, whereas the 3.9-[mu]m channel by itself only gives the cloud-top droplet effective radius. Better cloud phase detection will be possible at night using the ABI 8.5-[mu]m channel. Research designed to exploit these new sensor capabilities is underway.

TURBULENCE. Turbulence ["clear-air turbulence" ("CAT")] has long been considered a priority problem for commercial aircraft, and is the single greatest cause of injuries to flight attendants and passengers, as well as having the potential to cause structural damage to aircraft. Turbulence affecting aircraft is a small-scale phenomenon with length scales of tens to hundreds of meters. This makes it particularly difficult to observe with conventional observing systems, models, or satellites. It may be possible, however, to use satellite observations to identify larger-scale features or conditions that can lead to clear air or convectively induced turbulence. In this case, the satellite "products" or "interest fields" would need to be integrated with other observational systems and model output to produce an operationally useful warning product.

Mountain waves. One common source of turbulence is from breaking waves associated with flow over mountains or other terrain features. Wave clouds can often be observed in high-resolution satellite imagery, but the mere presence of wave structures is by itself not necessarily an indication of a turbulence hazard (Uhlenbrock et al. 2007). In addition, mountain waves do not always generate clouds. Wave structures in clear air can sometimes be observed in high- resolution water vapor imagery, where the wave motions result in humidity patterns that mimic the appearance of wave clouds (see Fig. 10). As part of ASAP, we are working on the development of an automated mountain wave identification algorithm for clouds and water vapor features that can be used with either GOES or MODIS imagery. Our current approach examines the water vapor channel for sudden changes in the magnitude and direction of the gradient by applying Laplacian-of-Gaussian (LoG) filtering. In our wave detection application of the LoG filter, the "edges" correspond to sudden changes in water vapor content. The sigma value used in the LoG filter affects the wavelength of the wave patterns enhanced by the LoG filter. Larger sigma values enhance larger wavelengths, while smaller sigma values enhance shorter wavelengths.

Changes in water vapor gradient direction are determined by finding the points in the LoG image where the sign changes between neighboring pixels. This "zero crossing" point represents a change in water vapor gradient direction that could be an indicator of turbulence.

Figure 10 shows an example using a MODIS image over an area in Colorado where a pilot reported turbulence within 30 min of the MODIS overpass. Wave patterns are clearly visible in the contrastenhanced 6.7-[mu]m channel shown in Fig. 10a. The zero crossing of the LoG image can be seen in Fig. 10b. The wave patterns are clearly detected. Brighter pixels represent a higher-magnitude gradient difference across the zero-crossing point, and therefore are more likely regions of turbulence.

The LoG method should be able to be used to quantify the strength and direction of mountain waves at different wavelength scales. By combining the LoG method with other image-processing and pattern recognition techniques, we hope to quantify the overall wave structure and find the relationship between the water vapor wave patterns and turbulence. This same technique may also be suitable for identifying gravity waves produced by convective clouds or systems, as modeled by Lane et al. (2003). In this case the wave detection capabilities would be used in conjunction with a generalized CCM to identify strong updrafts based on the cloud-top cooling rates. In combination with other observational systems and model output fields, this information may be able to help identify areas of convectively induced turbulence.

CAT within tropopause folds. Another example of a satellite- derived turbulence interest field is a product that identifies areas of tropopause folding, where the lower boundary of the stratosphere folds down into the troposphere. These folds frequently lead to dynamical instability that can result in areas of enhanced turbulence (Gidel and Shapiro 1979; Shapiro 1980). To this end, a turbulence prediction tool has been developed based on an upper- tropospheric specific humidity product that is derived from the GOES water vapor channel and corrected for satellite viewing angle and temperature (Moody et al. 1999; Wimmers and Moody 2001). The specific humidity product, termed the GOES layer average specific humidity (CLASH), varies closely with tropopause height. Strong gradients in the CLASH field have been found to be associated with tropopause folding (Wimmers and Moody 2004a,b). Figure 11 illustrates a CLASH field over the United States. Areas of yellow and red indicate subtropical air, while areas of blue and purple identify areas of dry polar air. Estimated areas of tropopause folding are represented as gray ribbons of uniform width, based on an assumed average size of tropopause folds. As a turbulence "interest field," this sort of product is integrated with model- generated fields to assign height limits for the folding, and to put this possible source of turbulence into a broader context. Using automated flight data records from commercial airlines, we have preliminarily found that light-or-greater turbulence occurs in a subset of these areas with a frequency of 25%. Research is underway to integrate the CLASH turbulence interest fields into national turbulence guidance systems.

Sounder simulation studies. In a longer-term effort, the ASAP science team is conducting an observing system simulation experiment (OSSE) to evaluate the potential for future hyperspectral sounding systems, such as are currently planned for the NPOESS Preparatory Project (NPP) and GOES-R (Schmit et al. 2005) satellites, to infer the presence of turbulence. In this experiment, NCAR scientists are generating three-dimensional output fields from very-highresolution NWP models that are capable of resolving the onset and development of turbulence on scales of importance to aircraft. These three- dimensional data fields will then be used by researchers at UW- CIMSS as "ground truth" for a simulation of the expected hyperspectral signals that will be observed by advanced high- spectral-resolution IR sounders, such as the Cross-track Infrared Sounder (CIS), Infrared Atmospheric Sounding Interferometer (IASI), and Hyperspectral Environmental Suite (HES) (see the appendix). All hyperspectral instruments will be capable of observing/measuring atmospheric features at unprecedented vertical resolution, with an ability to resolve small- or moderate-scale temperature or water vapor perturbations through evaluation of many channels. While the current GOES sounder footprint may be too large to detect turbulent structures directly (~8-10 km), these new hyperspectral instruments will have the capability to identify areas of high wind shear or high spatial variability in temperature and humidity and through the monitoring of the returned signals with time or by tracking the motion of larger-scale features that may be associated with turbulence. This information, when used in conjunction with other turbulence observations and diagnostics from NWP models, may become an important contribution to an integrated turbulenceforecasting system.

VOLCANIC ASH. Inadvertent penetrations of volcanic ash clouds have caused several nearly catastrophic commercial airline incidents, including engine failures from particulate ingestion and the total obscuration of cockpit windscreen from etching by the acidic aerosols (Miller and Casadewall 2000). The international community has established a network of Volcanic Ash Advisory Centers (VAACs) that are responsible for monitoring the airspace in their assigned geographical regions for hazardous eruptions and to issue warnings as needed to all aircraft in their respective areas (Romero 2004). At present, the VAACs make use of satellite observations, pilot reports, and alerts received from volcano observatories and ground-based monitoring systems. The FAAs oceanic weather PDT is working to assist the VAACs through improved detection and processing systems. In support of the FAA's contributions, the ASAP team is working with the PDT to develop and test advanced satellite- based volcanic ash detection techniques for higher reliability, earlier detection, fewer false alarms, and better identification of the plume height.

Current detect/on methods. While ash clouds are sometimes quite impressive in VIS satellite imagery, it is frequently difficult to distinguish ash clouds from natural, nonvolcanic water and ice clouds. Fortunately, suspended ash particles have some distinct spectral signatures that can be used to enhance their detection through multispectral image processing. The most commonly used multispectral technique is based on a strong 12-[mu]m absorption signature (e.g., Praia 1989), which causes a negative 11- minus 12- [mu]m T^sub B^ difference. This technique is often termed the reverse-absorption method, and has been the basis for a number of different detection techniques, as discussed by Ellrod et al. (2003). In addition to ash, volcanic eruptions usually release high concentrations of SO^sub 2^, which has strong absorption bands at 7.3 and 8.6 [mu]m. While the ash cloud and the SO^sub 2^ plume do not always follow the same downwind trajectories (e.g., Seftor et al. 1997), the presence of an SO^sub 2^ plume can help confirm a volcanic eruption and improve the reliability of complementary algorithms, such as those based on reverse absorption methods. At present, SO^sub 2^ absorption channels are available on the high- resolution Moderate Resolution Imaging Spectroratimeter (MODIS) imager and the Atmospheric Infrared Sounder (AIRS) sounding instrument in polar orbit on NASA's Earth Observing System (EOS) satellites, with the next-generation operational geostationary satellite (e.g., GOES-R) having similar capabilities.

Ongoing research. ASAP efforts to enhance volcanic ash detection are concentrated on improving satellitebased multispectral techniques for identifying and tracking ash clouds and for estimating their height. Once demonstrated and evaluated, these techniques will be transferred to the VAACs.

Volcanic eruptions and the resultant ash plumes can exhibit considerable variability in their structure and composition, with potentially variable responses to different detection algorithms. While individual detection techniques can often enhance the visibility of the ash plume, the resultant images are themselves somewhat noisy and prone to false detections. To minimize these problems we have concentrated on techniques that may be able to reduce the number of false alarms, either by the integration of results from multiple detection algorithms or through the inclusion of additional satellite channels that might provide a more robust detection capability. In much of our volcanic ash work we have emphasized the use of MODIS imagery as our primary test bed, because MODIS provides observations at most wavelengths of potential interest and very high spatial resolution.

ASAP ash detection studies are currently underway at UW-CIMSS and UAH. One approach being investigated is based on the standard reverseabsorption technique supplemented with VIS (0.65 [mu]m) and near-IR (3.75 [mu]m) channel data (Pavolonis et al. 2006). This new technique is physically based and globally applicable. Unlike the reverse-absorption technique, the new algorithm does not require the presence of a negative 11-^m minus 12-[mu]m T^sub B^ difference that is often absent, especially in tropical eruptions. Figure 12 shows the results of applying both the Pavolonis et al. algorithm and the standard reverse absorption technique to a 2004 eruption of Manam in Papua New Guinea. The new approach is able to identify ice clouds that are heavily contaminated with volcanic aerosols, even though these clouds generally lack a reverse-absorption signature. The reverse-absorption results only detect the portions of the low- level ash plume that are quite near the volcano, while having many small areas of negative T^sub B^ difference showing up in the nearby convective clouds. In contrast, the new algorithm produces a clean, false-alarm-free integrated product. For a simple implementation of the standard reverseabsorption technique (Pavolonis et al. 2006), the false-alarm rates are much improved, especially related to the cleanness of plume identification.

In a separate effort, the same statistical clustering approach used in the convective cloud mask, discussed in the "Convective initiation" section (T. A. Berendes et al. 2007, unpublished manuscript, hereafter BER), is applied to ash detection. The clustering algorithm makes use of MODIS channels that are known to be useful for detecting volcanic ash (0.65-, 0.55-, 3.75-[mu]m reflectance and 11.0-[mu]m T^sub B^). In addition, several difference channels (11-12, 3.75-11 [mu]m) were selected for inclusion based upon a principal component analysis. Finally, a ratio channel (3.7-[mu]m reflectance/0.65 [mu]m) was added to the cluster (Pavolonis et al. 2006).

The BER clustering algorithm was trained using several days of data over the Manam volcano, and 5 clusters out of approximately 160 were identified as containing volcanic ash. Figure 13 shows a true color MODIS image of the Manam volcano in the upper left. The clusters are shown upper right using a randomized color scheme. The lower-left image in Fig. 13 shows the threshold 3.7/0.65 [mu]m and the lower-right image shows the volcanic ash clusters in red. The volcanic ash clusters correspond well with the ash areas indicated by the 3.7/0.65-[mu]m image; one drawback is that ice clouds contaminated with ash are not well identified by this method.

This clustering technique has shown considerable potential for producing clean identification of prominent ash plumes over both land and water, but requires regional training and, like many techniques, has difficulty with optically thin ash plumes. Recent developments in the clustering algorithm have allowed for the creation of texture channels that can be used as features for training the clustering. Textural clustering may yield significant improvements in the detection of thin, dry ash, especially over land.

Identifying the height of an ash cloud can be very difficult, particularly when the ash is semitransparent. At operational VAAC centers, plume-height identification is performed by observing plume drift and correlating with global model wind fields at various altitudes. When the atmosphere exhibits directional shear, this can be a very accurate approach, but it cannot be relied upon for all situations.

Ash heights and timely detection. ASAP height identification efforts are based on the following two potential retrieval algorithms: 1) the CO2 ratio technique applied to MODIS data (Platnick et al. 2003), and 2) a "split window" (e.g., 11- and 12- [mu]m regions) onedimensional variational data assimilation (IDVAR) (e.g., Marks and Rodgers 1993; Heidinger et al. 2005). The MODIS CO2 ratio technique utilizes the four CO2 absorption channels on the MODIS (channels 33-36) in a conceptually consistent manner as described in the "Basic cloud properties" section. The splitwindow technique operates under the premise that for a known atmospheric state, the split-window measurements for a single-layer cloud are primarily determined by the following three factors: 1) cloud temperature, 2) cloud emissivity, and 3) cloud particle size and shape distribution. If an ash cloud is identified by the four- channel technique or some other method, a reasonable assumption about the particle size and shape characteristics can made, leaving only the cloud temperature and emissivity to be retrieved within the IDVAR system from the measured 11- and 12-[mu]m radiances. The resultant cloud temperature can then be matched to a pressure or height within an atmospheric sounding as is performed in the CO2 ratio technique.

Although the CO2-slicing method should be more accurate than the split-window method for midand upper-tropospheric plumes, many imagers do not (and will not) have CO2 absorption channels, so the split-window algorithm may be a useful substitute when CO2 slicing is not possible. For lowertropospheric ash plumes the split-window method should be superior. The error characteristics of each height retrieval algorithms will depend on the accuracy of the cloud particle size and shape distribution assumed by the algorithm. Current work is focused on quantifying the cloud-height errors as a function of cloud microphysics and optical thickness.

The use of MODIS imagery limits the likely operational use of many of these techniques to real-time demonstrations, based on the need for prompt availability of the data at the VAAC. With the launching of the GOES-R series of operational geostationary satellites, however, these same techniques should be readily transferred to full operational use.

No matter how reliable our ash identification techniques become, or how well we can track and predict the spread of the ash cloud, it is critical that the initial alerts and warnings be issued rapidly. At the 2004 Conference on Volcanic Ash and Aviation Safety in Alexandria, Virginia, for example, it was noted that volcanic ash can reach commercial flight levels in as little as 5 min. Even with extensive monitoring systems currently in place at key volcanoes and an international network of VAACs, it can take hours for eruptions in remote areas to be identified and evaluated, and for formal alerts to be issued.

In order to speed the initial detection of explosive eruptions, ASAP investigators are also examining untraditional approaches, such as using Department of Defense's (DoD's) IR surveillance satellites. These satellites are the backbone of the global missile early warning system and make observations of the Earth in the shortwave IR regions, with 10-s revisit rate (Pack et al. 2000). DoD scientists have documented the detection of volcanic eruptions in this imagery and have reported evidence for a characteristic spike in radiance coincident with the initiation of explosive eruptions (Moore and Rice 1984; Sparks et al. 1986; Pack et al. 2000). The attenuation of IR radiation by intervening cloud and water vapor places documented limitations on the detection of all eruptions.

Observations by DoD satellites, if processed routinely for volcanic eruptions and made available to the VAACs, may be a uniquely valuable enhancement to existing volcano-monitoring systems and would permit rapid analysis of eruptions that could prove to be a threat to aviation. SUMMARY. The fundamental objective of the ASAP program is to support the existing FAA weather Product Development Teams, encourage the full use of existing and future satellite- based observing systems for aviation applications, and accelerate the transition of research products to operational use.

Observations and products from research satellites, however, are not always appropriate for routine operational use. Data access can be slow, and one-of-a-kind satellites like Aqua and Terra are only overhead at certain times of the day. These satellites, however, can play an important role in validating current weather products with researchers gaining experience in the use of their advanced sensors in the process.

In many cases research sensors also serve as prototypes for future operational sensing systems. NASA's MODIS imager, for example, can be considered the prototype for both upcoming NPP and NPOESS Visible/Infrared Imagery Radiometer Suite (VIIRS) satellites, as well as the GOES-R Advanced Baseline Imager (ABI). Experience with the research instruments will translate directly to the accelerated use of the upcoming operational instruments with already tested products.

ACKNOWLEDGMENTS. Research support for the Advanced Satellite Aviation Weather Products (ASAP) initiative was provided by the NASA Applied Sciences Program under the direction of Ronald Birk and by the NASA Aviation Safety and Security Program's (AvSSP's) Aviation Weather Information element directed by Paul Stough. We also gratefully appreciate the assistance of John Haynes (NASA) on the ASAP initiative.

ASAP efforts are dependent on the coordination and cooperation of the Federal Aviation Administration's Aviation Weather Research Program (FAA's AWRP), under the direction of Ms. Gloria Kulesa. We also gratefully acknowledge the input and guidance we have received from the FAA's Product Development Teams, including the teams for in- flight icing (Marcia Politovich, NCAR), turbulence (Bob Sharman, NCAR), convective weather (Marilyn Wolfson, MIT-LL, and Cindy Mueller, NCAR), oceanic weather (Cathy Kessinger, NCR), and national ceiling and visibility (Paul Herzegh, NCAR). Discussions with Carl Rice and Dee Pack at the Aerospace Corporation are acknowledged.

We also want to thank two anonymous reviewers for their very helpful, constructive comments that have considerably improved the quality of this paper.

1 Future sensors planned for GOES-R and NPOESS platforms provide more channels than are currently available, allowing for potential improvements to cloud property retrieval algorithms. Increased spatial and temporal resolutions will also enhance the utility of satellite cloud products for aircraft-icing operational systems.

REFERENCES

Ackerman, S. A., 1996: Global satellite observations of negative brightness temperature differences between 11 and 6.7 [mu]m. J. Atmos. Sci., 53, 2803-2812.

Bedka, K. M., and J. R. Mecikalski, 2005: Application of satellite-derived atmospheric motion vectors for estimating mesoscale flows. J. Appl. Meteor., 44, 1761-1772.

Bedka, S. T., W. F. Feltz, A. T. Schreiner, and R. E. Holz, 2007: Satellite derived cloud top pressure product validation: Using aircraft based cloud physics lidar data from the AtREC field campaign. Int. J. Remote Sens., in press.

Bankert, R. L., 1994: Cloud classification of AVHRR imagery in maritime regions using a probabilistic neural network. J. Appl. Meteor., 33, 909-918.

Bernstein, B., F. McDonough, M. Politovich, B. Brown, T. Ratvasky, D. Miller, C. Wolff, and G. Cunning, 2005: Current Icing Potential (CIP): Algorithm description and comparison with aircraft observations. J. Appl. Meteor., 44, 118-132.

Cai, H., and Coauthors, 2006: Enhancements of NCAR Auto-Nowcast System using NRL, ASAP, MM5 and TAMDAR data. Preprints, 12th Conf. on Aviation Range and Aerospace Meteorology, Atlanta, GA, Amer. Meteor. Soc., CD-ROM, P11.1.

Donovan, M. F., E. R. Williams, C. Kessinger, G. Blackburn, P. H. Herzegh, R. L. Bankert, S. Miller, and F. R. Mosher, 2007: The identification and verification of hazardous convective cells over oceans using visible and infrared satellite observations. J. Appl. Meteor. Climatol, in press.

Ellrod, G. P., and J. Nelson, 1996: The use of GOES-8 multispectral imagery for the detection of aircraft icing regions. Preprints, Eighth Conf. on Satellite Meteorology and Oceanography, Atlanta, GA, Amer. Meteor. Soc., 168-171.

____, B. H. Connell, and D. W. Hillger, 2003: Improved detection of airborne volcanic ash using multispectral infrared satellite data. J. Geophys. Res., 108, 4356 doi:10.1029/2002JD002802.

Fujita, T. T., 1985: The downburst. Satellite and Mesometeorology Research Project, University of Chicago, Research Paper 210, 122 pp. [NTIS PB-148880.]

____, 1986: DFW microburst. Satellite and Mesometeorology Research Project, University of Chicago, Research Paper 217, 155 pp.

____, and H. R. Byers, 1977: Spearhead echo and downburst in the crash of an airliner. Mon. Wea. Rev., 105, 129-146.

____, and F. Caracena, 1977: An analysis of three weather- related aircraft accidents. Bull. Amer. Meteor. Soc., 58,1164-1181.

Gidel, L. T., and M. A. Shapiro, 1979: The role of clear air turbulence in the production of potential vorticity in the vicinity of upper tropospheric jet stream-frontal systems. J. Atmos. Sci., 36, 2125-2138.

Glickman, T., Ed., 2000: Glossary of Meteorology. 2d ed. Amer. Meteor. Soc., 855 pp.

Haggerty, J. A., and Coauthors, 2005: Integration of advanced satellite cloud products into an icing nowcasting system. World Weather Research Programme Symp. on Nowcasting and Very Short Range Forecasting, Toulouse, France, WWRP/WMO.

Heidinger, A. K., M. D. Goldberg, D. Tarpley, A. J. Jelenak, and M. J. Pavolonis, 2005: A new AVHRR cloud climatology. Applications with Weather Satellites II, Vol. 5658, Bellingham, WA, SPIE, 197- 205.

Hobbs, P. V., and A. Rangno, 1985: Ice particle concentrations in clouds. J. Atmos. Sci., 23, 2523-2549.

Isaac, G., and Coauthors, 2005: First results from the Alliance Icing Research Study II. Proc. 43d AIAA Aerospace Sciences Meeting, Reno, NV, AIAA 20050252.

Kaplan, M. L., Y.-L. Lin, A. J. Riordan, K. T. Waight, K. M. Lux, and A. W. Huffman, 1999: Flight safety characterization studies, Part I: Turbulence categorization analyses. Interim Subcontractor Report to Research Triangle Institute, NASA Contract NASI99074.

Khaiyer, M., P. Minnis, B. Lin, W. L. Smith Jr., and A. D. Rapp, 2003: Validation of satellite-derived liquid water paths using ARM SGP microwave radiometers. Proc. 13th ARM Science Team Meeting, Broomfield, CO, ARM.

Lane, T. P., R. D. Sharman, T. L. Clark, and H. M. Hsu, 2003: An investigation of turbulence generation mechanisms above deep convection. J. Atmos. Sci., 60, 1297-1321.

Levine, M. D., and S. I., Shaheen, 1981: A modular computer vision system for picture segmentation and interpretation. IEEE Trans. Pattern Anal. Machine Intelligence, Vol. PAMI-3, 540-556.

Marks, C. J., and C. D. Rodgers, 1993: A retrieval method for atmospheric composition from limb emission measurements. J. Geophys. Res., 98, 14 939-14 952.

Mecikalski, J. R., and K. M. Bedka, 2006: Forecasting convective initiation by monitoring the evolution of moving cumulus in daytime GOES imagery. Mon. Wea. Rev., 134,49-78.

____, and Coauthors, 2002: NASA Advanced Satellite Aviation- weather Products (ASAP) Study Report. NASA Tech. Rep., 65 pp.

Miller, T. P., and T. J. Casadevall, 2000: Volcanic ash hazards to aviation. Encyclopedia of Volcanoes, H. Sigurdsson, Ed., Academic Press, 915-930.

Minnis, P., D. P. Kratz, J. A. Coakley Jr., M. D. King, D. Garber, S. Mayor, D. R Young, and R. Arduini, 1995: Cloud optical property retrieval (subsystem 4.3). Vol. 3, Cloud Analyses and Radiance Inversions (Subsystem 4) (subsystem 4.3), Clouds and the Earth's Radiant Energy System (CERES) Algorithm Theoretical Basis Document, NASA RP 1376,135-176.

____, and Coauthors, 2001: A near real time method for deriving cloud and radiation properties from satellites for weather and climate studies. Proc. 11th Conf. Satellite Meteorology and Oceanography, Madison, WI, Amer. Meteor. Soc., 477-480.

Moody, J. L., A. J. Wimmers, and J. C. Davenport, 1999: Remotely sensed specific humidity: Development of a derived product from the GOES Imager channel 3. Geophys. Res. Lett, 26, 59-62.

Moore, J. G., and C. J. Rice, 1984: Chronology and character of the May 18, 1980 explosive eruptions at Mt St. Helens. Special Volume on Explosive Volcanism, National Academy of Sciences, 133- 142.

Mueller, C., T. Saxen, R. Roberts, J. Wilson, T. Betancourt, S. Dettling, N. Oien, and J. Yee, 2003: NCAR AutoNowcast System. Wea. Forecasting, 18, 545-561.

Murray, J. J., 2002: Aviation weather applications of Earth Science Enterprise data. Earth Observing Magazine, Vol. 11, No. 8, 26-30.

Nair, U. S., R. C. Weger, K. S. Kuo, and R. M. Welch, 1998: Clustering, randomness, and regularity in cloud fields 5. The nature of regular cumulus cloud fields. J. Geophys. Res., 103, 11 363-11 380.

National Research Council, 1980: Technological and Scientific Opportunities for Improved Weather and Hydrological Services in the Coming Decade. Select Committee on the National Weather Service, National Academy Press, 87 pp.

____, 1991: Toward a New National Weather Service: A First Report. National Weather Service Modernization Committee, National Academy Press, 67 pp.

____, 1994: Weather for Those Who Fly. National Weather Service Modernization Committee, National Academy Press, 100 pp.

Pack, D. W., C. J. Rice, B. J. Tressel, C. J. Lee-Wagner, and E. M. Oshika, 2000: Civilian uses of surveillance satellites. Crosslink Magazine, The Aerospace Corporation, Vol. 1, No. 1, 2-8. [Available online at www.aero. org/publications/crosslink/pdfs/V1N1.pdf.]

Pavolonis, M. J., W. F. Feltz, A. K. Heidinger, and G. M. Gallina, 2006: A daytime complement to the reverse absorption technique for improved automated detection of volcanic ash. J. Atmos. Oceanic Technol, 23, 1422-1444. Platnick, S., M. D. King, S. A. Ackerman, W. P. Menzel, B. A. Baum, J. C. Riedi, and R. A. Frey, 2003: The MODIS cloud products: Algorithms and examples from Terra. IEE Trans. Geosci. Remote Sens., 41, 459-473.

Prata, A. J., 1989: Observations of volcanic ash clouds in the 10- 12 micrometer window using AVHRR/2 data. Int. J. Remote Sens., 10, 751-761.

Rauber, R. M., and A. Tokay, 1991: An explanation for the existence of supercooled water at the top of cold clouds. J. Atmos. Sci., 48,1005-1023.

Roberts, R. D., and S. Rutledge, 2003: Nowcasting storm initiation and growth using GOES-8 and WSR-88D data. Wea. Forecasting, 18, 562-584.

____, T. Saxen, C. Mueller, J. Wilson, A. Crook, J. Sun, and S. Henry, 1999: Operational application and use of NCAR's thunderstorm nowcasting system. Preprints, Int. Conf. on Interactive Information and Processing Systems, Dallas, TX, Amer. Meteor. Soc., 158-161.

Romero, R., 2004: The international airways volcano watch (IAVW). Proc. Second Int. Conf. on Volcanic Ash and Aviation Safety, Alexandria, VA, Office of the Federal Coordinator for Meteorological Services and Supporting Research, 1-10.

Schmit, T. I., M. M. Gunshor, W. P. Menzel, J. J. Gurka, J. Li, and A. S. Bachmeier, 2005: Introducing the next-generation advanced baseline imager on GOESR. Bull. Amer. Meteor. Soc., 86, 1079-1096.

Schreiner, A. J., and T. J. Schmit, 2001: Derived cloud products from the GOES-M imager. Preprints, llth Conf. on Satellite Meteorology and Oceanography, Madison, WI, Amer. Meteor. Soc., 420- 423.

____, D. A. Unger, W. P. Menzel, G. P. Ellrod, K. I. Strabala, and J. L. Pellett, 1993: A comparison of ground and satellite observations of cloud cover. Bull. Amer. Meteor. Soc., 74, 1851- 1861.

Seftor, C. J., N. C. Hsu, J. R. Herman, P. K. Bhartia, O. Torres, W. I. Rose, D. J. Schneider, and N. Krotkov, 1997: Detection of volcanic ash clouds from Nimbus7/TOMS. J. Geophys. Res., 102, 16 749- 16 760.

Shapiro, M. A., 1980: Turbulent mixing within tropopause folds as a mechanism for the exchange of chemical constituents between the stratosphere and troposphere. J. Atmos. Sci., 37, 994-1004.

Sparks, R. S. J., J. G. Moore, and C. J. Rice, 1986: The initial giant umbrella cloud of the May 18th 1980 explosive eruption of Mount St. Helens. J. Volcanol. Geothermal Res., 28, 257-274.

Tag, P. M., R. L. Bankert, and L. R. Brody, 2000: An AVHRR multiple cloud-type classification package. J. Appl. Meteor., 39, 125-134.

Trepte, Q., Y. Chen, S. Sun-Mack, P. Minnis, D. Young, B. Baum, and P. Heck, 1999: Scene identification for the CERES cloud analysis subsystem. Proc. 10th Conf. on Atmospheric Radiation, Madison, WI, Amer. Meteor. Soc., 169-172.

Triplett, W., 2005: The calculators of calm. Vol. 19, No. 6, Smithsonian Air & Space Museum, 54-61.

Uhlenbrock, N. L., K. M. Bedka, W. F. Feltz, and S. A. Ackerman, 2007: Mountain wave signatures in MODIS 6.7-um imagery and their relation to pilot reports of turbulence. Wea. Forecasting, 22, 662- 670.

Velden, C. S., C. M. Hayden, S. J. Nieman, W. P. Menzel, S. Wanzong, and J. S. Goerss, 1997: Upper-tropospheric winds derived from geostationary satellite water vapor observations. Bull. Amer. Meteor. Soc., 78, 173-195.

____, T. L. Olander, and S. Wanzong, 1998: The impact of multispectral GOES-8 wind information on Atlantic tropical cyclone track forecasts in 1995. Part I: Dataset methodology, description, and case analysis. Mon. Wea. Rev., 126, 1202-1218.

____, and Coauthors, 2005: Recent innovations in deriving tropospheric winds from meteorological satellites. Bull. Amer. Meteor. Soc., 86,205-223.

Vivekanandan, J., G. Thompson, and T. Lee, 1996: Aircraft icing detection using satellite data and weather forecast model results. Proc. FAA Int. Conf. on Aircraft Icing, Springfield, VA, FAA, 365- 373.

Wimmers, A. J., and J. L. Moody 2001: A fixed-layer estimation of upper tropospheric specific humidity from the GOES water vapor channel: Parameterization and validation of the altered brightness temperature product. J. Geophys. Res., 106 (D15), 17 115-17 132.

____, and ____, 2004a: Tropopause folding at satellitebased spatial gradients. I. Verification of an empirical relationship. J. Geophys. Res., 109, D19306, doi:10.1029/2003JD004145.

____, and ____, 2004b: Tropopause folding at satelliteobserved spatial gradients. II. Development of an empirical model. J. Geophys. Res., 109, D19307, doi:10.1029/2003JD004146.

Wolff, C., B. Bernstein, and F. McDonough, 2005: Nowcasting aircraft icing conditions using GOESderived satellite products. WWRO Symp. on Nowcasting and Very Short Range Forecasting, Toulouse, France, WWRO/WMO.

Wylie, D. P., and W. P. Menzel, 1989: Two years of cloud cover statistics using VAS. J. Climate, 2, 380-392.

AFFILIATIONS: MECIKALSKI AND BERENDES-Atmospheric Science Department, University of Alabama in Huntsville, Huntsville, Alabama; FELTZ, K.M. BEDKA, S. T. BEDKA, AND WIMMERS-Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin-Madison, Madison, Wisconsin; MURRAY AND MINNIS-Chemistry and Dynamics Branch, NASA Langley Research Center, Hampton, Virginia; JOHNSON, HAGGERTY, AND BERNSTEIN-Research Applications Laboratory, NCAR, Boulder, Colorado; PAVOLONIS-Advanced Satellite Products Branch NOAA/ NESDIS. Madison, Wisconsin; WILLIAMS-Lincoln Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts

CORRESPONDING AUTHOR: John R. Mecikalski, National Space Science and Technology Center, University of Alabama in Huntsville, 320 Sparkman Drive, Huntsville, AL 35805

E-mail: john.mecikalski@nsstc.uah.edu

The abstract for this article can be found in this issue, following the table of contents.

DOI:10.117S/BAMS-88-10-1589

In final form 17 May 2007

(c)2007 American Meteorological Society

APPENDIX: GLOSSARY OF NEW SATELLITE AND SATELLITE-BORNE OBSERVING SYSTEMS.

This appendix is intended to update Glickman (2000) to reflect new satellites and instruments that are not in

Source: Bulletin of the American Meteorological Society

More News in this Category