Advertisement

Pure and Applied Geophysics

, Volume 176, Issue 5, pp 2017–2043 | Cite as

Applications of Geostationary Satellite Data to Aviation

  • Gary P. EllrodEmail author
  • Kenneth Pryor
Article

Abstract

Weather is by far the most important factor in air traffic delays in the United States’ National Airspace System (NAS) according to the Federal Aviation Administration (FAA). Geostationary satellites have been an effective tool for the monitoring of meteorological conditions that affect aviation operations since the launch of the first Synchronous Meteorological Satellite (SMS) in the United States in 1974. This paper will review the global use of geostationary satellites in support of aviation weather since their inception, with an emphasis on the latest generation of satellites, such as Geostationary Operational Environmental Satellite (GOES)-R (16) with its Advanced Baseline Imager (ABI) and Geostationary Lightning Mapper (GLM). Specific applications discussed in this paper include monitoring of convective storms and their associated hazards, fog and low stratus, turbulence, volcanic hazards, and aircraft icing.

Keywords

Aviation weather geostationary satellites GOES-R Himiwari Meteosat Fengyun ABI AHI SEVIRI thunderstorms convective initiation fog low stratus volcanic ash volcanic SO2 aircraft icing microbursts geostationary lightning mapper 

1 Introduction

Weather is the most important factor in air traffic delays in the United States’ National Airspace System (NAS), accounting for more than one half of all delays lasting > 15 min (Federal Aviation Administration 2017). Geostationary (hereafter, GEO) satellites have been an effective tool for the monitoring of meteorological conditions that affect aviation operations since the launch of the operational Synchronous Meteorological Satellite (SMS) in the United States in 1974 (and the first of the subsequent series designated as the National Oceanic and Atmospheric Administration’s (NOAA) Geostationary Operational Environmental Satellite (GOES)-A launched in 1975). (A series of experimental Applications Technology Satellites (ATS) had been launched by the National Aeronautics and Space Administration (NASA) between 1966 and 1974 to serve as a proof of concept.) The advantage of GEO satellites is that they frequently view the same large area of the earth’s atmosphere from their equatorial vantage point, providing temporal continuity in the observation of evolving storm systems and other environmental hazards over land and ocean areas. There is an excellent coverage of tropical and mid-latitude weather from GEO satellites, but the imagery and derived products are degraded in Polar Regions due to the oblique viewing angle, requiring augmentation with polar-orbiting spacecraft such as the U.S. NOAA and the European Organization for the Exploitation of Meteorological Satellite (EUMETSAT) agency’s MetOp series.

Prior to the GEO observations from space, aviation weather conditions and hazards could only be monitored using the surface and upper air observation network, land-based weather radar, and sporadic pilot reports (PIREPS). GEO satellites helped to create a more complete picture of ongoing weather throughout the airspace system. This paper will summarize the use of GEO observations for aviation applications from the earliest basic imaging spacecraft to the more advanced technology such as GOES-R.

2 Brief History of Geostationary Satellite Technology

Internationally, SMS and GOES in the U. S. were followed by the European METEOSAT and Japan’s Geostationary Meteorological Satellite (GMS) in 1977. Coverage of the Indian sub-continent began with the Indian National Satellite (INSAT)-1B launched in 1983. China’s Fengyun (FY) series of satellites began with FY-2 in 1997. The early satellites (SMS, GOES) had a simple two-band imager known as the Visible and Infrared Spin Scan Radiometer (VISSR) (Suomi and Krauss 1978) that viewed the earth in 2-km resolution visible (0.5 µm), and 8-km thermal infrared (IR 10–11 µm) spectral bands. Since the satellites were spin-stabilized, they only viewed the earth for 5% of each rotation, which reduced the potential for rapid scan imaging. The visible band was used for cloud, surface, and aerosol imaging during daytime. The lower resolution IR band observed clouds day and night but could not see aerosols unless they were exceedingly dense.

A water vapor (WV) absorption band centered near 6.5 µm was first introduced on the METEOSAT-1 launched in 1977 (Morel et al. 1978). Images from the WV band were able to observe changes in moisture patterns in the middle and upper troposphere associated with features such as jet streams, gravity waves, and upper troughs even in the absence of clouds, which was not previously possible. Beginning in the 1980s, WV band imagery became standard on all GEO satellites.

Each successive generation of GEO satellites has resulted in significant upgrades to their capabilities. A 12-band sounder, the VISSR Atmospheric Sounder (VAS), was introduced on GOES-4 in 1981. The GOES I-M series first launched in 1994 (Menzel and Purdom 1994) had a greatly improved, 19-band VAS with significant improvements in IR resolution from 7 or 14 km (depending on the detector used) to 4 and 8 km. A five-channel imager incorporated into GOES I-M (designated GOES 8–12 once in operational status) was capable of frequent (every 5–15 min), simultaneous observations at 4-km resolution in short-wave IR (3.9 µm), WV IR (6.7 µm), long-wave window (11.0 µm) and “split-window” IR (12.0 µm) bands for the first time. The three-axis stabilized GOES I-M series was capable of more efficient imaging and better quality signal to noise ratios (Menzel et al. 1998).

GOES-VAS also produced vertical thermal and moisture profiles and horizontal pressure gradient winds for input to numerical weather prediction (NWP) models and microburst prediction, as well as 8-km resolution-derived image products applicable to aviation, such as Lifted Index (LI) or cloud top pressure (CTP).

In the Eastern Asia field-of-view, Korea launched its first GEO satellite [Communication, Ocean and Meteorological Satellite (COMS)] in 2010 (Kim and Ahn 2014), with an imager that is similar to GOES-VAS and Japan’s Multi-functional Transport Satellite (MTSAT) series of satellites.

While this paper will review aviation applications of prior GEO spacecraft, it will focus on advanced capabilities of the newest generations of GEO satellites: the GOES-R series in the U. S. (Schmit et al. 2005, 2017), METEOSAT Second Generation (MSG) in Europe (Schmetz et al. 2002), the advanced Himiwari in Japan (Bessho et al. 2016), and the FY-4 series in China (Yang et al. 2017). The MSG was launched in 2002 and became operational by 2004, while Himiwari and GOES-16 became operational in 2015 and December 2017, respectively. The first of China’s advanced FY-4 series of GEO satellites was launched in late 2016, with several follow-on spacecraft planned through 2025. The METEOSAT Third Generation (MTG), which will be similar to GOES-R, is planned for launch by late 2021 (Bézy et al. 2005; Mohr 2014). Both MTG and FY-4 have advanced multispectral sounders onboard.

3 GOES-16 Sensors and Applications

3.1 Advanced Baseline Imager (ABI)

The Advanced Baseline Imager (ABI) on GOES-16 is a state-of-the-art instrument that is based on capabilities of several previous spacecraft including: prior GOES imagers and sounders, the Visible Infrared Imaging Radiometer Suite (VIIRS) (Hillger et al. 2013) on the polar-orbiting Suomi National Polar-orbiting Partnership (NPP), the High Resolution Infrared Sounder (HIRS) on the NOAA polar-orbiting series, and the Moderate Resolution Imaging Spectroradiometer (MODIS) onboard the NASA polar-orbiting Terra and Aqua spacecraft (Schmit et al. 2005, 2017).

Table 1 shows the spectral bands on the advanced geostationary satellites, their central wavelength (µm), nominal satellite instantaneous geometric field-of-view (IGFOV, km), sample aviation uses, and availability of each band on ABI, the Advanced Himiwari Imager (AHI), MSG Spinning Enhanced Visible and Infrared Imager (SEVIRI) and the FY-4 Advanced Geostationary Radiation Imagery (AGRI). The AHI is identical to the ABI except for the inclusion of a “green” visible band at 0.51 µm, which, along with the “blue” (0.47 µm) and “red” (0.64 µm) bands, allows the creation of “true color” imagery for easier interpretation of surface features (e.g., Miller et al. 2016). A simulated true color product is created using ABI by incorporating the “blue” and “red” visible image data with corrected “green” channel data from AHI to create a lookup table (Miller et al. 2016). Both the ABI and AHI have IGFOVs of 0.5 km for the “red” visible band (0.64 µm), 1.0 km for the “blue” visible band (0.47 µm) and two near-IR bands, and 2 km for two additional near-IR and ten IR bands.
Table 1

Description of spectral bands on the advanced geostationary satellites including:, central wavelength, instantaneous geometric field-of-view (IGFOV) (km), uses in aviation forecasting and hazard detection, and availability on the GOES-16 ABI, Himiwari AHI and MSG SEVIRI

Slight differences may exist in wavelengths of equivalent bands between the spacecraft

aFY-4 Midwave IR spans the 6.9–7.3 µm wavelengths

The current MSG SEVIRI features 12 spectral bands with resolutions of 1 km for visible and 3 km for IR. The MTG series will have advanced capabilities, including a 16-band Flexible Combined Imager (FCI) similar to ABI and AHI, and a Lightning Imager (LI) on one platform, and separate IR and UltraViolet (UV) sounders on another spacecraft (Bézy et al. 2005; Mohr 2014). Specific applications of some of the spectral bands will be described in more detail in later sections.

In addition to more spectral bands and higher resolution, GOES ABI will be able to sample the earth at a much higher rate (Schmit et al. 2017). The full disk image, previously only sampled once every 3 h, will now be possible every 15 min. The Continental U. S. scan (and its equivalent for the west GOES spacecraft) will be 5-min interval, and two so-called mesoscale sectors (about 1000 × 1000 km) will be available once per minute (Schmit et al. 2015). This is especially important for aviation weather support, to detect rapidly changing conditions and provide timely warnings to aircraft.

Even more significantly, level 1b radiances will be directly assimilated into numerical weather prediction (NWP) models and used for model verification (Schmit et al. 2017). In addition, Schmit et al. (2008) suggest that data from the Cross-track Infrared Sounder (CrIS) and other polar-orbiting IR sounders in conjunction with the finer spatial resolution ABI data will provide a useful substitute for current sounder information for radiance uses within NWP. In effect, the improvements in spatial and temporal resolution of ABI radiances should positively impact NWP model data assimilation through the process of noise and error reduction.

3.2 Geostationary Lightning Mapper (GLM)

A Geostationary Lightning Mapper (GLM) was incorporated into GOES-R, the first time that an instrument of this type was placed into geostationary orbit. The GLM is based on the NASA Lightning Imaging Sensor (LIS), previously flown on the polar-orbiting Tropical Rainfall Measurement Mission (TRMM), and the Optical Transient Detector (OTD), flown on an independent low earth orbit spacecraft (Goodman et al. 2013). Lightning mappers are also planned for MTG (Bézy et al. 2005; Mohr 2014) and FY-4 (Yang et al. 2017).

The GLM has a sub-point resolution of 8 km, degrading to 14 km near the Earth’s limb. The sensor unit observes each pixel in the field of view for a period of 2 ms (ms) at a visible wavelength of 777 nm (0.777 µm), which is an oxygen emission band. A lightning flash detection efficiency (DE) (or probability of detection) of 70% is required by users and it was estimated that an actual DE of close to 90% may be attainable. Detection efficiency [DE (%)] is defined as:
$${\text{DE }}\left( \% \right) \, = \, N_{\det } / \, N \, \times \, 100$$
(1)
where N is the total number of lightning flashes, and Ndet is the number of flashes detected by the instrument.
The false alarm rate (FAR) required for the GLM was 5%, where FAR is defined as:
$${\text{FAR }}\left( \% \right) \, = \, N_{\text{f}} / \, N \, \times \, 100$$
(2)
where Nf is the number of false detects by the instrument. FARs may be caused by noise in the data or sun glint. Since lightning flashes may be observed in adjacent pixels, groups of 2 × 2 pixels (an area of 16 × 16 km) are used to produce a “flash”.

The GLM currently cannot distinguish between in-cloud (IC) and cloud to ground (CG) lightning, so it returns total lightning at each location. The main uses of GLM in aviation will be to help identify the most intense areas of convection in oceanic regions where surface-based radar data are sparse, or to help diagnose potential severe storms over both land and sea. Research using ground-based lightning detection systems has discovered that rapid increases in total lightning activity (referred to as lightning “jumps”) often precede severe weather from thunderstorms (Schultz et al. 2009).

3.3 Space Weather Instruments

Among the suite of sun-pointing sensors on GOES-R are two that are used to monitor activity from solar storms that could disrupt terrestrial power grids and communication systems. Of particular importance to aviation are the Extreme Ultraviolet and X-Ray Irradiance Sensor (EXIS). Both EUV and X-ray spikes from solar flares bombard the ionosphere and an excess can lead to high-frequency (HF) blackouts that disrupt communication between aircraft and ground controllers. In the U.S., the NOAA Space Weather Prediction Center (SWPC) issues warnings of radio blackouts at low latitudes. There is also the risk of radiation exposure for aircrews that frequently fly on polar routes. Aircrews are now considered “radiation workers” by European and international agencies.

4 Specific Applications of Geostationary Satellite Data to Aviation

4.1 Convective Initiation

Organized, deep thunderstorms can be extremely disruptive to aircraft by producing enroute hazards such as hail and strong turbulence, and by forcing flights to reroute, leading to late arrivals, ground delays and airport gridlock. Identification of developing thunderstorms, especially near approaches to major airline hubs, is critical in managing air traffic flow and reducing costly inflight diversions. The earlier that developing storms are identified, the earlier controllers can make plans to accommodate traffic volume. For short duration flights, 2–6 h advanced notice is needed, with long haul flights requiring 4–8 h (Federal Aviation Administration 2017).

High-resolution visible GOES imagery, combined with IR bands can help anticipate developing storms and update short-range convective forecasts. Subjectively, when convection develops in a destabilizing air mass, there is a rapid increase in the area and brightness of cumulus clouds in visible images, along with a rapid cooling in IR imagery. Initiation of convective storms was observed to occur in GOES images along sea breeze fronts associated with land–water interfaces, or with mergers and intersections of cumulus cloud lines which may precede the first radar echoes by several hours (Purdom 1976). Once storms have formed, the leading edge of the storm-initiated mesohigh (associated with the surface gust front) is often marked by a narrow line of cumulus known as an arc line (Purdom 1973). When arc lines merge or intersect other unstable regions, the resulting surface moisture convergence may initiate new convection which can become severe (Purdom 1976).

An objective approach to predicting convective initiation (CI) (defined as cumulus cells that reach a radar reflectivity of ≥ 35 dBZ) was developed in the early 2000s (Mecikalski and Bedka 2006) using GOES-11 and GOES-12 data. It was later improved in a system known as Satellite Convection Analysis and Tracking System, version 2 (SATCASTv2) (Walker et al. 2012) using GOES-13. While partially successful based on evaluations in four regions east of the Rockies (combined Probability of Detection (POD) was 72% (with a high of 85% in the central U.S.) for cloud objects that were tracked), evaluations indicated that false alarm rates (FAR) were unacceptably high (48–60%). [A CI system based partially on SATCAST was also developed in Europe using data from MSG SEVIRI (Merk and Zinner 2013)]. The under-detection and false alarms for the SATCAST CI were mainly due to: (1) a mismatch of cloud object tracking and (2) sub-resolution initial cloud development that led to false indication of rapid cloud growth, and (3) cirrus cloud contamination leading to undetected CI. Also, rapid initial growth of cumulus clouds observed by satellite can be misleading, since atmospheric conditions such as a mid-layer capping inversion may not be favorable for continued development into convective storms. As expected, the performance of CI was worse at night due to IR-only data, and there were frequent instances of cloud objects being lost in the transition from daytime visible/IR combination to only IR at night.

A GOES-R version of the CI system was significantly improved when the satellite-only version was enhanced by numerical weather prediction (NWP) data (Mecikalski et al. 2015). A total of 25 predictors were used, nine of which were satellite-based (Table 2). Two enhanced regression techniques, logistic regression (LR) and random forest (RF) were used to predict the likelihood of CI for each successive time interval. The FAR for both methods was lowered to 25–35%. A similar system has been developed in Korea based on the COMS data, with similar results (Han et al. 2015).
Table 2

List of satellite parameters used in the GOES-R convective initiation (CI) algorithm (still under evaluation, specific values may change in future versions)

Satellite CI predictors

10.7 µm Tb (K)

13.3–10.7 µm Tb (K)

6.5–10.7 µm Tb (K)

15-min 10.7-µm cloud top cooling rate (K)

15-min change of 13.3–10.7 µm Tb (K)

Dominant cloud type at time 2

Dominant cloud type at time 1

Object size at time 1 (1 km pixels)

Object size at time 2 (1 km pixels)

The GOES-R algorithm can be summarized as follows: first of all, a cloud “mask” is produced that identifies only cumulus clouds within an area of interest. The techniques involved in creating the cloud mask use visible brightness gradients, visible brightness thresholding, 10.7 µm IR brightness temperature (Tb) time-differences to screen out immature cumulus and non-cumulus clouds, and textural analysis of visible brightness values to remove low, thick stratus clouds from being incorrectly classified as cumulus. Then, low-level atmospheric motion vectors (AMVs) were determined to predict where each cumulus cloud would be located in successive images. Finally, nine satellite parameters and 16 NWP parameters are used for each time interval. The satellite parameters (shown in Table 2) are based on IR Tb values, trends of Tb, and multi-spectral Tb differences. (The original satellite-only CI algorithm had eight threshold tests). The NWP variables were derived from the 13-km rapid update (RAP) models. The output is a probabilistic value from 0 to 1 for the cloud object to reach the CI state, in which ground-based radar reflectivity reached ≥ 35 dBZ. Note that the GOES CI is not an operational “baseline” product, thus the procedures used to create CI such as the predictors shown in Table 2 may be further modified.

The CI “nowcasting” system is in experimental use by forecasters at the U.S. FAA, National Weather Service (NWS) Forecast Offices, the Aviation Weather Center (AWC) Testbed, and the NOAA Hazardous Weather Testbed (Mecikalski et al. 2015). It is expected that the current version of CI can provide advanced warning of developing convective storms with greater than 50% probability in the 30–50 min range on average. Recent examples of the use of CI available to operational forecasters can be seen in Gravelle et al. 2016.

The advantages of GOES-16 ABI imagery with the CI system are that both IR and visible data will have greater resolution, and the images will be more frequent (maximum 1–5 min versus the current 5–15 min). The additional spectral bands will also help in the identification of cloud top glaciation and cloud top heights.

4.2 Locating Hazardous Thunderstorms Over Land and Oceanic Areas

4.2.1 Global Convective Diagnostic (GCD)

Long overwater flights often traverse multi-layered mid-latitude frontal systems with embedded convection or tropical weather systems near the Inter-Tropical Convergence Zone (ITCZ). Avoidance of thunderstorm activity in these systems is a priority to prevent encounters with severe turbulence or icing. For pre-flight planning, international high-level Significant Weather (SIGWX) charts are issued by World Area Forecast Centers (WAFC) in London and Washington, DC, and for inflight avoidance Significant Meteorological (SIGMET) advisories are issued by numerous Meteorological Watch Offices (MWOs) and transmitted to airlines and other end-users.

Identification of oceanic deep convection uses primarily geostationary satellite imagery (IR and visible) and lightning data along with a few isolated island radar stations. Coverage using satellite data is excellent with the current network of Himiwari, GOES, and MSG. However, it is sometimes difficult to identify regions of the most active storm cells within multi-layered cold cloud systems.

Subjective methods to identify strong convection in cold cloud systems with satellite IR include the use of a threshold cloud top temperature (CTT) (i.e. 215 K) or identification of CTT gradients. An objective method using the Tb difference between 10.7 µm IR and 6.7 µm water vapor (WV) bands on GOES-12 was developed by Mosher (2001, 2002) for use at the U. S. NWS Aviation Weather Center (AWC) in Kansas City. The Global Convective Diagnostic (GCD) flagged a cloud as a possible thunderstorm if the WV–IR Tb difference was 1 °C or less. Verification of the GCD using storm height measurements with the Precipitation Radar on the Tropical Rainfall Measuring Mission (TRMM) spacecraft (Martin et al. 2008) showed that GCD significantly over-predicted the coverage of deep convection. However, performance improved when the Brightness Temperature Difference (BTD) was lowered to near 0 °C (no specific value was recommended). Also, GCD significantly out-performed the use of a single IR threshold of 215 K, except for a single case involving a Great Plains squall line.

The GCD is now used in combination with the Naval Research Laboratory’s (NRL) satellite-derived Cloud Top Height (CTOP) (Donovan et al. 2008) and Cloud Classification (CC) algorithms (Tag et al. 2000), and the Earth Networks global ground-based lightning strike data (Liu and Heckman 2011) by the University Center for Atmospheric Research (UCAR) Research Applications Laboratory (RAL) to form a product known as the Convective Diagnostic Oceanic (CDO) product. A CDO index value (0–6) is created using a weighted average of the CTH, GCD, GOES-R Overshooting Tops (OT) (Bedka et al. 2010), and the Earth Networks lightning accumulation algorithm. The experimental CDO is available online at: http://www.rap.ucar.edu/projects/ocn/realtime_sys/wxHazards/. An example is shown in Fig. 1.
Fig. 1

Convective Diagnosis Oceanic (CDO) product derived from GOES-EAST (16) and GOES-WEST (13) valid 0000 UTC, 5 January 2018. Cloud top heights are shown by the color scale at right as flight levels (100′s of feet). Storm intensity is shown by color scale in legend at the top. Only CDO values of 2 (moderate storm) through 5 (extreme storm) are displayed

The description and evaluation of CDO has been completed and published by Donovan et al. (2008). The use of instruments on the Tropical Rainfall Measuring Mission (TRMM) to: (1) observe lightning (lightning imaging system), (2) precipitation intensity (precipitation radar), and (3) cloud properties [visible and infrared radiometer system (VIRS)] determined that both convective clouds with lightning (TRW) and without lightning (CLW) could be detected with a POD as high as 90% when lightning is used as a criterion for “hazardous” convective clouds. There is an ongoing effort to use these products to create the Ensemble Prediction of Oceanic Convective Hazards (EPOCH) system (Bass 2017). EPOCH real-time output of convective forecasts from 12 to 48 h in the future is viewed at the AWC but is not available to the public.

4.2.2 Lightning Activity with Oceanic Convection

Another important tool for thunderstorm identification in remote areas is lightning data from both time of arrival (TOA) and direction finding (DF), ground-based lightning systems first developed in the late 20th century. These systems observe electromagnetic pulses emitted by lightning with detection efficiency (DE%) of about 90% but with coverage limited to mostly continental areas. A World Wide Lightning Location Network (WWLLN) introduced in the early 2000s (Lay et al. 2004) uses the emission of very low-frequency (VLF) (1–24 kHz) radio pulses from lightning that can travel many thousands of km to expand detection coverage to oceanic areas. Location accuracy with WWLLN is good (15 km) but DE is estimated to be very low (10% for CG, 6% for total lightning). The GLM thus offers an excellent observational tool for frequent monitoring of oceanic convection in support of aviation with its near 90% DE and low FAR (Goodman et al. 2013).

An example of how GLM can identify oceanic thunderstorm activity associated with Hurricane Harvey over the western Gulf of Mexico is shown in Fig. 2. Parallax-corrected GLM lightning flashes for a 5-min period prior to a GOES visible image at 1317 UTC on 25 August 2017 is color-coded, with red being the oldest and yellow the newest lightning. Note the absence of lightning near the center of Harvey and westward toward the Texas coast. Lightning near the center of tropical cyclones (< 100 km) is typically episodic (DeMaria et al. 2012). Little or no lightning usually indicates a steady-state storm, while large lightning densities occur with either rapidly strengthening or weakening storms (DeMaria et al. 2012). Since overshooting cloud tops can be seen near the eye, they nevertheless represent a hazard to high-altitude flights through this area.
Fig. 2

A GOES-16 ABI Band 1 (visible) image at 1317 UTC on 25 August 2017 that shows Hurricane Harvey near the Texas coast, overlaid by GLM lightning data (non-operational, still undergoing beta-testing at the time of this writing) for a 5-min period prior to the image time. Lightning data are color-coded according to time (red is oldest, yellow is most recent). Note that the GLM data were parallax-corrected while the GOES-16 imagery was not

(Source: Lindstrom 2017)

4.2.3 Severe Storm Identification

The identification of potentially severe convective storms over land has traditionally involved measurements of the time rate of change of CTT (K) or the growth (expansion of the area of the cirrus anvil) using both visible and IR data (Sikdar et al. 1970; Adler and Fenn 1979). Based on analysis of 5-min rapid scan data from Synchronous Meteorological Satellite (SMS)-2, storms with colder IR tops that grew more rapidly were more likely to be severe. Mean vertical velocity profiles for the severe elements were twice those of non-severe cases (Adler and Fenn 1979). Reports of tornadic activity on the ground occurred during, or immediately following, a rapid expansion of cold cloud top areas, indicating rapid ascent of the cloud. A lead time of as much as 30 min was observed. More recent work using GOES data, such as the University of Wisconsin Cloud Top Cooling (UW-CTC) algorithm (Hartung et al. 2013), shows good correlation between high rates of CTC and strong radar reflectivity (as measured by the WSR-88D in the Central U. S.), vertically integrated liquid (VIL) and maximum estimated size of hail (MESH). Lead times of 0–60 min were observed.

Features observed in individual IR images may also indicate the presence of a strong or severe convective storm. One of these is the “overshooting top” (OT) in which single or multi-pixel clusters of CTTs are significantly colder than the surrounding anvil cirrus. An OT by itself is most often associated with heavy rainfall (Negri and Adler 1981: Dworak et al. 2012). An example of the ability of the GOES-R (16) ABI to better detect OTs is shown in Fig. 3 for a Mesoscale Convective System (MCS) in northern Wisconsin on 16 May 2017. This feature sometimes has a V-shape, with warmer IR temperatures within the V downshear from the cold core (McCann 1983; Heymsfield and Blackmer 1988). The temperature difference in the cold-warm “couplet” varied from 7 to 17 K for severe storms. The cause of the downstream warm spot is believed to be subsidence of stratospheric air over the cloud top into the cirrus anvil, perhaps as a wave-like feature. A median lead time of 30 min was observed from identification of the enhanced-V to the occurrence of severe weather, although the low probability of detection (~ 40%) indicated that many severe storms do not have this feature (McCann 1983).
Fig. 3

A comparison of GOES-13 IR (top) versus a non-operational GOES-16 IR mage (bottom) for strong storms over northern Wisconsin, USA, on 16 May 2017. The GOES-16 image is approximately 5 min later. Overshooting tops are the dark red to black circular clusters of pixels with temperatures near − 70 °C (203 K)

(Source: Bachmeier 2017a)

An objective, satellite-based OT detection algorithm has been developed for GOES-R using 5–30 min 11 µm IR data from GOES-12 for a 6-year period from January 2004 to December 2009 (Dworak et al. 2012). The best results (false alarms ~ 15%) occurred when minimum CTT was < 200 K and when the satellite was in rapid scan (5 min interval imaging). This emphasized the importance of high-resolution, high-frequency imaging to observe rapidly evolving cloud top features. However, the GOES-R OT algorithm is not operational in the baseline product suite.

Dark bands observed along the upstream edge of strong convective storms in GOES 6.7-µm WV imagery are also good indicators of possible severe weather (Ellrod 1990). The feature is likely caused by subsidence as strong upper level winds encounter a convective cloud with strong updrafts that act as a barrier to the flow. Of 147 incidences of this feature during the spring and early summer of 1989, 82% of the storms were severe. The predominant type of severe weather reported (83%) when this feature was present was large hail (≥ 3/4″ diameter). The occurrence of the upstream dark band was most often observed in the central and northern Plains of the U. S.

The addition of the GLM on GOES provides a powerful additional tool for monitoring en route convection for severity. Rapid increases in total lightning flashes that often precede severe weather activity, known as lightning “jumps,” have been observed in data from land-based Lightning Mapper Array (LMA) systems (Schultz et al. 2009). An algorithm that determined whether total lightning flash rates exceeded the average for non-severe storms by two standard deviations (2σ) or more performed the best, with POD of 87% and acceptable FARs (23–33%). This algorithm and others are in the process of evaluation for use with the GOES-R series.

4.2.4 Convective microbursts

The use of data from the GOES sounders for the nowcasting of convective storm potential commenced in the 1980s with the application of the VISSR Atmospheric Sounder (VAS) (Smith et al. 1981) to complement conventional surface and upper level observations. The VAS was first applied specifically to downburst potential assessment for the 2 August 1985 Dallas-Ft. Worth (DFW), Texas microburst storm (Ellrod 1989), which led to the crash of a commercial jetliner, killing 133. The following parameters were found to be most effective for prediction of the microburst that occurred with this case: large temperature lapse rates (°C km−1 from 850 to 700 hPa), high-convective available potential energy (CAPE) (J kg−1), large vertical equivalent potential temperature (θe) difference (850–700 hPa), and low mid-tropospheric relative humidity (RH%). The existing GOES sounder-derived microburst products were designed to diagnose risk based on favorable environmental thermodynamic profiles for severe convective storm development. Menzel et al. (1998) describes the role, performance, and weather forecasting applications of the most recent generation of GOES sounders, as well as the thermodynamic profile generation process. (Although the GOES-R (16) instrument suite does not include a sounder, vertical temperature and humidity profiles can be generated with the ABI and high-resolution numerical models using methods described by Lee et al. (2014).)

The typical horizontal scale of a single-cell convective storm and larger downburst (or macroburst) is near 10 km (Byers and Braham 1949). Considering the 10-km spacing of sounding retrievals, the GOES-VAS was well-suited to observe horizontal variations in environmental conditions and associated parameters that indicate risk of strong winds produced by downbursts. In addition, a new version of the sounding physical retrieval algorithm (Li et al. 2008) was implemented into GOES operations in 2011.

Pryor and Ellrod (2004a, b) outlined the development of a suite of GOES sounder-derived products to assess the presence of conditions favorable for dry and wet microbursts. That study introduced the wet microburst severity index (WMSI) product to calculate potential magnitude of convective downbursts in humid environments over the eastern United States. The WMSI incorporated (surface-based) convective available potential energy (CAPE) as well as the vertical equivalent potential temperature difference (Δθe) between the surface and mid-troposphere (Atkins and Wakimoto 1991).

The Microburst Windspeed Potential Index (MWPI) (Pryor 2010, 2012, 2014), compared to the WMSI, was designed to quantify the most relevant factors in convective downburst generation in intermediate thermodynamic environments by incorporating: (1) surface-based CAPE, (2) temperature lapse rate between the 670 and 850 hPa levels (|Г|), and (3) differences in dewpoint-temperature depression (ΔDD) between the 670 and 850 hPa levels. The MWPI was incorporated into a predictive linear model developed in the manner exemplified in Caracena and Flueck (1988). The MWPI formula consists of a set of predictor variables (i.e. dewpoint depression, temperature lapse rate) that generates output of expected microburst risk. Analysis of microbursts during the Joint Airport Weather Studies (JAWS) project (Wakimoto 1985; Caracena and Flueck 1988) identified the following favorable environmental characteristics for high-plain dry microbursts: (1) low-surface dewpoint temperatures, (2) high-convective cloud base, (3) small mid-tropospheric ΔDD, and (4) high sub-cloud lapse rate. Consideration of 50 downburst events over Oklahoma and Texas during the summer of 2009 revealed that for the majority of downburst events, CAPE was greater than 1000 J kg−1 and ΔDD was greater than 5 °C. Srivastava (1987) also noted that a minimum lapse rate of 5 °C km−1, associated with heavy precipitation and a high reflectivity factor (> 50 dBZ), was necessary for intense downdraft generation. Thus, the scaling factors of 1000 J kg−1, 5 °C km−1, and 5 °C, respectively, are applied to the MWPI algorithm to yield a non-dimensional MWPI value that expresses wind gust potential on a scale from one to five (Pryor 2015):
$$\begin{aligned} {\rm{MWPI }} & \equiv {\rm{ }}\left\{ {\left( {{\rm{CAPE}}/1000{\rm{ J k}}{{\rm{g}}^{ - 1}}} \right)} \right\}{\rm{ }} \\ & \quad + {\rm{ }}\left\{ {\Gamma /5\;^\circ {\rm C} \, {\rm{ k}}{{\rm{m}}^{ - 1}} + {\rm{ }}\left( {{{\left( {T - {T_{\rm{d}}}} \right)}_{850}} - {{\left( {T - {T_{\rm{d}}}} \right)}_{670}}} \right)/5\;^\circ {\rm{C}}} \right\} \end{aligned}$$
(3)

The MWPI algorithm is expected to be most effective in assessing downburst wind gust potential associated with ordinary cell and multi-cell convective storms in weak wind shear environments. Employing 13-km resolution rapid refresh (RAP) model data as a proxy for GOES-16 ABI vertical sounding profile datasets, the MWPI and associated wind gust potential was effectively displayed for a severe downburst event over Idaho in June 2017. The reader is referred to Pryor (2017) for a detailed discussion of the convective environment and evolution of this downburst event.

More recently, the launch of GOES-16 and subsequent activation of the ABI has resulted in the increased accessibility to high spatial and temporal satellite image datasets for convective storm monitoring and nowcasting. Traditional meteorological satellite techniques for deep convective storm monitoring, including the WV–IR BTD, (Schmetz et al. (1997), have recently been extended by multi-band techniques for diagnosing attributes of a favorable downburst environment. The WV-IR BTD has demonstrated the ability to infer convective storm structure signatures (i.e. overshooting tops, lateral dry-air notches) that signify the occurrence potentially damaging straight-line winds resulting from downbursts, and, with 2-km resolution GOES-16 ABI data, the BTD product is able to detect fine detail signatures more effectively when compared to 4-km resolution GOES-13-15 imagery (Pryor 2017). Accordingly, a three-band BTD algorithm incorporating split window IR bands (11, 12 μm) and a WV band (7.3 μm) produces an output BTD that is proportional to downburst potential and supplements the sounder-derived MWPI. Figure 4 demonstrates the ability of two ABI-derived BTD products in diagnosing favorable storm structural patterns and environmental conditions for downburst occurrence in the upper Great Basin region on 4 June 2017. High thunderstorm-induced winds up to 28 ms−1 (54 kt) were observed at the Idaho National Laboratory (shown by an ‘*’ in the images), where three-band BTD values of 50–60 K indicated wind gust potential of 26–31 ms−1 (50–60 kt).
Fig. 4

a GOES-16 ABI three-band BTD product image (based on bands 10 (7.3 µm), 14 (11.2 µm), and 15 (12.3 µm) at 2100 UTC, 4 June 2017 and b GOES-16 WV-IR BTD product image at 2255 UTC, 4 June 2017. An ‘asterisk’ marks the location of the Idaho National Laboratory, where a wind gust of 54 knots was recorded. Black arrows in b mark the location of dry-air notches

(from Pryor 2017)

4.3 Fog and Low Cloud Detection

4.3.1 Area coverage

Observing low clouds and fog with geostationary satellite imagery initially involved subjective pattern recognition that required considerable skill by the analyst/forecaster. The distinctive smooth, bright appearance of low clouds in daytime visible imagery makes them relatively easy to identify. At night, the problem became more difficult due to the formation of fog and low clouds beneath thermal inversions. In IR imagery, the tops of these clouds can be warmer, the same temperature, or even cooler than surface temperatures depending on the depth of the thermally mixed boundary layer.

Based on a bi-spectral approach used with AVHRR image data (Eyre et al. 1984), GOES VAS imagery was able to provide better detection of fog at night (Ellrod et al. 1989) over the Continental U. S. With the advent of multi-spectral imager bands at 30-min intervals on GOES-8 in 1995, the use of a bi-spectral BTD between the short-wave IR (SWIR 3.9 µm) and window IR (10.7 µm) bands was shown to be far more effective in observing fog and stratus at night than single IR bands alone (e.g., Ellrod 1995; Lee et al. 1997). Image animations of the “fog product” could observe development, spread, and even dissipation of disruptive low clouds. The physical basis of the technique is that the radiative emissivity in the SWIR is lower than that of the LWIR, thus resulting in detection of cooler sub-cloud conditions with SWIR, and a BTD of ~ 1 to 4 °C. Figure 5 shows an area of low clouds and cirrus in Oklahoma in IR, and SWIR bands, along with a temperature comparison in the two IR bands along the line A–B.
Fig. 5

Comparison of IR Tb values across the transect A–B in CH2 (3.9 µm) and CH4 (10.7 µm) on 24 March 1999 showing the basis for nighttime fog imagery using GOES

One of the shortcomings of earlier geostationary systems is that the limited resolution (4–8 km) in the IR bands resulted in poor detection of valley fog. The GOES-R ABI and Himiwari AHI provide further upgrades for this product, with 2-km IR resolution and faster repeat scans (1–5 min). Figure 6 shows a comparison of the GOES-16 fog product versus one from GOES-13 for the same area of the Eastern U.S. on 17 October 2017. The improved quality of GOES-16 is evident due to the improved resolution and better precision (lower instrument noise).
Fig. 6

Nighttime fog imagery for GOES-13 (left) versus the same product for GOES-16 (right) at 0615 UTC (0215 EDT) on 17 October 2017. Valley fog is shown as yellow or cyan, cloud-free areas as gray, and cirrus is black

(Source: University of Wisconsin-CIMSS 2017)

4.3.2 Estimating Time of Clearing

The time required for the clearing or lifting of fog and low clouds is an important factor in aviation forecasting. Certain airports (especially San Francisco, Seattle, and Los Angeles in the U. S.) are especially prone to weather delays caused by fog. Early attempts to forecast dissipation of fog using GOES were based on the digital value of the brightness difference between the fog and adjacent cloud-free areas, a measure of fog thickness (Gurka 1978). While this method was effective, it could not be used until about 1 h after sunrise.

An IR enhancement technique developed using GOES sounder data from 1989 to 1992 based on SWIR-IR BTD values (Ellrod 1995) resulted in pre-dawn estimates of fog or stratus thickness which could then be used to determine approximate fog dissipation time after sunrise. Verification of this method using higher resolution GOES IR imager data from 1997 to 2001 and aircraft cloud top reports revealed a correlation of 0.625.

A case that demonstrated the potential improvements in forecasting the time of clearing of low stratus for airport operations that can be achieved using GOES-16 has been documented at San Francisco International Airport (SFO) (Eckert 2017). On the morning of March 3, 2017, a 3-h long Ground Delay Program (GDP) was initiated due to developing stratus near SFO. Based on GOES-16 time lapse visible imagery at 5-min intervals, clearing was observed to occur more rapidly than with the operational GOES-West images at 15-min intervals. This resulted in the lifting of the GDP about an hour earlier than possible using legacy GOES-West images alone, leading to a net cost saving of $50 K to the airlines and passengers.

4.3.3 Estimating areas of IFR conditions

One of the shortcomings of satellite data in the detection of low clouds and fog is the uncertainty of the cloud base heights to determine the presence of instrument flight rule (IFR) (ceilings < 1000 ft, visibilities < 3 statute miles) or marginal visual flight rule (MVFR) conditions. There are pattern recognition techniques that provide clues, such as the texture of the cloud tops in IR and visible data, and cloud shadows. A low cloud base (LCB) product that combines the IR fog imagery with the surface temperature from ground observation sites was obtained to estimate areas where IFR conditions may exist (Ellrod 2002; Ellrod and Gultepe 2007). If the GOES IR cloud top minus surface shelter temperatures were ≤ 3 K and a low cloud was present in the GOES product, IFR ceilings (< 1000 ft) were considered likely. This product was verified for all regions of the U.S. between May and August 2000 in comparison with > 1500 Meteorological Aeronautical Reports (METARs) (Ellrod 2002). Cases where the low clouds were obscured by higher cloud layers were not verified. The POD ranged from a low of 58% in the Northeast to a high of 77% in the upper Midwest. The FAR varied between 15% in the Northeast to 4% in the South Central U.S. A modification of this scheme used operational numerical model data over southern Ontario, Canada, to remove the effects of mid and high clouds to produce better results (Gultepe et al. 2007).

GOES bi-spectral IR fog and low cloud imagery has been combined with surface temperatures and Rapid Refresh (RAP) model data to obtain probabilities of IFR (ceilings 500–1000 ft), low IFR (ceilings < 500 ft), and Marginal Visual Flight Rule (MVFR) (ceilings 1000–3000 ft) (Calvert and Pavolonis 2011). (At this time, there is no capability to provide information on surface visibilities from satellites.) Fog depth (km) is also determined. The requirements for these products are a fog/low clouds detection rate of 70%, and a cloud layer thickness accuracy of 500 m. The model data help to provide details on sub-cloud moisture and temperature conditions that reduce false alarms and improve detection in areas where cirrus obscures lower cloud layers. Instead of BTD, the GOES-R algorithms use a parameter called pseudo-emissivity (EMS) for the 3.9-µm data. EMS is the ratio of the observed 3.9-µm radiance (numerator) and the 3.9-µm blackbody radiance calculated using the 11-µm brightness temperature (denominator). The advantage of using EMS over BTD is that it results in better skill scores because it is less sensitive to scene temperature. These products have been generated experimentally and began using GOES-R (16) data in late 2017.

4.4 Aircraft Turbulence

4.4.1 Upper Level Clear Air Turbulence (CAT)

Clear air turbulence (CAT) occurs at high altitudes (6–15 km) in a nearly cloud-free atmosphere, and is usually associated with vertical wind shears near the jet stream and upper level fronts. It is one of the most common causes of accidents involving large commercial jet airliners, resulting in injuries to air crews and passengers (e.g. NTSB).

In satellite imagery, CAT has been observed in the vicinity of cirrus or moisture boundaries associated with the jet stream and frontal zones. The nature of the cirrus itself often has clues regarding the presence of CAT in the form of striations or bands that are oriented transverse to the flow (e.g. Knox et al. 2010). Wider, thicker bands have been associated with a higher risk of moderate–severe turbulence. These bands have also been observed on the periphery of mesoscale convective systems (MCS) and tropical cyclones, especially where they interact with mid-latitude or sub-tropical jet streams.

Transverse cirrus bands tend to be irregular in shape and orientation and can be seen in both IR and visible data. Evenly-spaced wave clouds known as “billows” that are caused by gravity waves and Kelvin–Helmholtz Instability (KHI) have typically been observed only in high-resolution visible imagery, but modern imagers such as AHI and ABI can be used to observe them in the IR and WV bands also. A high-pass filtering technique has been able to identify small-scale gravity waves associated with aircraft turbulence over the western Pacific Ocean using 2-km resolution WV imagery from AHI (Wimmers et al. 2018).

On the poleward side of the jet stream, the air is often clear, but 6.7-µm IR water vapor (WV) images can indicate the presence of CAT if there is progressive warming with time (usually displayed as darkening gray shades). An evaluation of more than 100 cases of WV darkening from 1983 to 1985 indicated that moderate or greater CAT was present 80% of the time (Ellrod 1985). The image darkening has been associated with strong middle and upper tropospheric subsidence that occurs in conjunction with frontogenetic processes and possibly intrusion of stratospheric air via tropopause folds.

An objective technique has been developed to identify tropopause folds using the strength of gradients in GOES Layer Average Specific Humidity (GLASH) derived from 6.7-µm WV data (Wimmers and Moody 2001, 2004a, b). The likelihood of CAT with the tropopause folds is dependent on the orientation, slope and width of the dark bands. Empirical rules have been developed to identify those tropopause folds that are likely to contain CAT (Wimmers and Feltz 2010). Verification using GOES WV imagery and automated aircraft Eddy Dissipation Rate (EDR) data was completed from Nov 2005 to Dec 2007. An accuracy of 53% for detection of moderate or greater CAT was achieved (the product requirement is at least 50%). The Tropopause Folding Turbulence Product (TFTP) is planned for future implementation with the GOES-R series.

4.4.2 Mountain waves

When moderate to strong winds blow across mountain ridges in the presence of a stable layer near the mountain tops, mountain waves occur. When moisture is present near the base of the inversion, wave cloud appearing as washboard-like patterns can be observed in satellite imagery. Even the earliest GOES imagery clearly showed these cloud patterns (Conover 1964). The wavelength of the clouds is proportional to the wind speed and instability (Fritz 1965). Wavelengths have been observed to increase during the day as diurnal heating proceeds. The longer wavelengths (observable in GOES IR imagery) are more likely to be associated with significant low-level turbulence than the shorter wavelengths (Ellrod 1985). Figure 7 shows mountain waves in the Japanese Himiwari AHI and Multi-Transport Satellite (MTSAT) WV images on 7 July 2015. Their detection was made possible by the higher resolution (2 km) of the AHI WV band.
Fig. 7

Mountain wave clouds observed in 4-km resolution 7.0-µm WV imagery from Japan’s MTSAT (top) and Himiwari AHI 2-km 6.9-µm WV (bottom) on 7 July 2015 at around 0330 UTC

(UW-CIMSS Satellite Blog)

Evidence of mountain waves at high altitudes is indicated by standing lee wave clouds (sometimes referred to as “banner clouds”) composed of cirrus that extend as far as hundreds of kilometers downwind from the mountain ridges. The cirrus plumes persist since the subsidence of the standing wave is not sufficient to dissipate the ice crystals. There is usually no turbulence in these plumes unless a narrow, stationary clear zone in the cirrus occurs in IR or WV images parallel to and just east of the lee slopes of the mountain ridges (Ellrod 1987). This feature is likely a result of a standing mountain wave, such as that described by Lilly (1978) that extends up into the lower stratosphere. A pronounced subsidence zone occurs just east of the mountain ridges leading to the dark zone in the WV images. Strong downslope windstorms often occur simultaneously with this feature along the east slopes of the Front Range in Colorado.

Figure 8 shows an example of such features in a GOES-16 Band 10 (7.2 µm) image associated with a strong mountain wave and downslope high-wind episode in Colorado and Wyoming on 29 December 2017. Frequent PIREPs of moderate to severe turbulence and mountain wave activity were obtained from aircraft departing or arriving at Denver International Airport (DEN) throughout the day. Additionally, surface winds gusted to more than 50 mph in Northern Colorado and southeast Wyoming.
Fig. 8

GOES-16 Band 10 (7.3 µm) water vapor imagery at a 1157 UTC and b 1802 UTC on 29 December 2017. Warm bands associated with katabatic downslope flow can be seen at (1) near the Wind River and Teton Ranges in Wyoming and at (2) near the Front Range in Colorado or the Medicine Bow range in Wyoming. Cold cirrus plumes extending downwind indicate possible mid-high-altitude mountain waves and turbulence

4.5 Volcanic Hazards Monitoring

Volcanic ash is extremely hazardous to jet aircraft that inadvertently fly through it. The ash is composed primarily of silicate particles that melt when ingested into the combustion chamber of a jet engine, causing severe loss of engine performance and perhaps, a complete shutdown of the engine. Although no known aircraft have been lost due to volcanic ash encounters, there were two well-known incidents that resulted in near-catastrophes to large jet airliners in the late 1980s and early 1990s (e.g. Casadevall 1994). These incidents lead to the formation of the Volcanic Ash Advisory Centers (VAACs) under the International Civil Aeronautical Organization (ICAO) of the World Meteorological Organization (WMO). Since visual sightings of volcanic ash clouds from the ground or from aircraft are limited, remote sensing techniques based on satellite and radar observations are critical to the VAACs and Meteorological Watch Offices (MWOs) in monitoring hazardous ash clouds and issuing advisories such as Volcanic Ash Advisories (VAAs) and SIGMETs to aircraft in flight. It has been estimated that volcanic ash can be present in air routes at altitudes greater than 9 km (30,000 ft) on approximately 20 days per year worldwide (Miller and Casadevall 2000).

Volcanic eruptions have been monitored with GOES satellite images since the late 1970’s and 1980’s using long-wave window IR and visible images. Powerful eruptions such as Mt. St. Helens (May 1980) and El Chichon, Mexico (March 1982) were observed in this way. Although single-band images are effective for a few hours after an eruption, as the ash clouds became thinner, it becomes difficult to distinguish them from thin cirrus clouds or the underlying surface. By the late 1980’s, it was learned that the temperature difference in two long-wave IR bands could be used effectively in observing thin ash clouds, initially with AVHRR (Prata 1989), and later with GOES (Rose and Mayberry 2000). Known as the “two-band split-window” (TBSW) or “reverse absorption” technique, it has become a benchmark for observing ash globally, and for measuring the effectiveness of improved multi-spectral techniques.

While the TBSW method is often effective, there are a number of deficiencies that have been identified (Hufford et al. 2000) such as (1) the inability to detect ash when the eruption cloud is opaque (usually within several hours of the eruption), (2) obscuration by ice embedded in the eruption cloud, and (3) the masking of low-level ash clouds by moist tropical atmospheres.

Ash clouds have also been observed by weather radar in a number of instances (e.g. Rose et al. 1995). However, the number of radars near active volcanoes is quite small, and their range is normally limited to about 300 km. There is also the risk of mistaking a thunderstorm for a volcanic eruption, especially in tropical regions. The example in Fig. 9 shows how satellite and radar can complement each other in observing and verifying volcanic eruptions. The Anatahan Volcano in the Marianas Islands in the Western Pacific Ocean erupted strongly on 5 April 2005, producing a dense ash cloud that rose to 15.5 km (51,000 ft). Even at more than 300 km from the location of the radar on Guam, strong reflectivity (50 dBZ) was observed. Due to relatively stable atmospheric conditions, there were no deep convective clouds in the region on this day. The TBSW satellite product from GOES-9 (right panel) showed a strong ash signal (white) around the edge of the high-altitude eruption cloud as well as a lower level ash plume caused by earlier activity. However, the central eruption cloud does not show the classic ash signature in the TBSW image product due to the opaque cloud column consisting of dense water and ice, in addition to ash.
Fig. 9

Comparison of radar reflectivity (0.5° scan) from NWS Guam WSR-88D radar (left), GOES-9 band 4 10.7-μm IR (center), and band 5 minus band 4 IR split window (right) at around 1725 UTC, 5 April 2005, following an eruption of Anatahan volcano

(Ellrod 2005)

Later improvements in ash detection were achieved by supplementing the split window product with the GOES 3.9-µm (SWIR) IR band (Ellrod et al. 2003). The 3.9-µm band is sensitive during the daytime to small cloud particles, resulting in increased reflectance from ash plumes hours after an eruption. Even at night, some improvements in detection were possible. By incorporating both visible and SWIR imagery with the TBSW (Pavolonis et al. 2005), it was shown that daytime volcanic ash detection could be significantly improved, with nearly one hundred times fewer false alarms globally than with just the use of the TBSW. Developed using multi-spectral data from Aqua MODIS, an algorithm to automatically detect volcanic eruptions using the four-band approach has been developed and has been implemented on GOES-16 ABI and possibly other platforms. Products available with this suite include: ash cloud height, ash loading (integrated mass), ash/dust effective radius, and ash probability, and true color volcanic plume composition (ash vs SO2).

In an effort to better utilize satellite data for volcanic cloud applications, NOAA, in collaboration with the Cooperative Institute for Meteorological Satellite Studies (CIMSS), has developed the VOLcanic Cloud Analysis Toolkit (VOLCAT). VOLCAT ingests large volumes of data, from many different satellites, to automatically detect volcanic eruptions and send SMS text messages and/or email alerts to relevant users when an eruption is detected. VOLCAT also automatically tracks and characterizes volcanic ash clouds in an effort to improve volcanic ash cloud forecasts (e.g. Chai et al. 2017). The VOLCAT algorithms utilize unique spectral, spatial, and temporal metrics within a supervised learning framework (Pavolonis 2010; Pavolonis et al. 2013, 2015a, b) to automatically detect volcanic ash clouds with an accuracy that is comparable to a skilled human analyst. Figure 10 shows an example of a VOLCAT RGB image derived from Himiwari-8 AHI depicting the ash plume from Kambalny Volcano, Russia, drifting south–southwestward over the northwest Pacific Ocean at 1030 UTC on 25 March 2017. The densest portion of the ash plume is outlined by the brown contour. The yellow color of the plume indicates the presence of both volcanic ash and SO2 (a combination of input from the red and green color guns representing contributions from the 12–11-µm TBSW and 11–8.5-µm band differences, respectively). The green lines are contrails from commercial jet aircraft that passed through the region earlier, as detected by the 11–8.5-µm IR difference which shows cloud phase.
Fig. 10

An RGB multi-spectral image from the VOLcanic Cloud Analysis Toolkit (VOLCAT) system on 1030 UTC, 25 March 2017 showing the ash plume from Kambalny Volcano, Kamchatka Peninsula, Russia. See text for more explanation

(Bachmeier 2017b)

The use of hyperspectral instruments such as the advanced infrared radiometric sounder (AIRS) or infrared atmospheric sounding interferometer (IASI) provides considerable benefits in both detection of volcanic ash and reduction of false alarms due to airborne dust and surface noise. One such example is a correlation method demonstrated with IASI data (Clarisse et al. 2010) that was significantly better than using benchmark BTD thresholds alone. Current and future deployment of such instruments in GEO orbits (such as the Geostationary Interferometric Infrared Sounder (GIIRS) on FY-4) provides a great promise for improved global ash detection.

For operational, real-time detection of erupting volcanoes and timely notification for aircraft in adjacent airspace, a response time of 5 min or less is desirable (Hufford et al. 2000). Even with GEO satellites, this has not been possible in the past. However, the improved rapid scan capabilities of the GOES-R ABI and Himiwari AHI imagers makes this requirement potentially achievable with 1-min mesoscale imaging frequency for GOES-16 ABI (2.5-min scanning for AHI), provided that a volcano suspected of imminent eruption is being properly monitored. It has been shown that rapid cooling of volcanic clouds during an eruption as observed with AHI 11-µm IR window band is significantly larger (by > 10 standard deviations) than the growth of convective clouds determined from a large database, providing confidence in their detection. An example occurred with the eruption of Kambalny Volcano on the Kamchatka Peninsula of eastern Russia on 25 March 2017. Based on Himiwari AHI 2.5 min scans, the growth rate observed was more than 13 standard deviations above the mean, signaling a likely eruption (Pavolonis 2018).

In addition to the detection of ash, remote sensing of sulfur dioxide (SO2) emitted from volcanoes can be important to help estimate climatic effects, as well as to warn aircraft in flight about volcanic activity. However, since the SO2 cloud is sometimes at a higher altitude than the ash, its movement can be in a completely different direction, depending on existing wind shears. Ultraviolet (UV) detectors originally designed to observe total ozone concentrations (such as NASA’s Total Ozone Mapping Spectrometer or TOMS) have been adapted to observe volcanic total column SO2 (Seftor et al. 1997). They rely on backscattering from SO2 in a UV band which differs enough from the ozone signature to allow discrimination between the two. Certain IR bands previously available on the GOES Sounder and MODIS (especially a water vapor absorption band centered at 7.3 µm) and now the GOES-R ABI and the Himiwari AHI are sensitive to SO2 absorption (Prata et al. 2004; Ackerman et al. 2008). Figure 11 shows the similarity between a two-band difference image (based on 7.3–9.7 µm bands) from the GOES Sounder with a TOMS SO2 Index product showing coverage of high-altitude SO2 from Soufriere Hills Volcano, Montserrat, on 13 July 2003 (Ellrod and Schreiner 2004).
Fig. 11

Image of GOES-8 Sounder band 9 (9.7 µm) minus band 10 (7.4 µm) at 1320 UTC, 13 July 2003 (left), versus NASA TOMS SO2 Index at approximately 1530 UTC

(From Ellrod and Schreiner 2004)

5 Aircraft Icing

Aircraft icing is a significant hazard to aircraft leading to loss of performance, especially among smaller general aviation and commuter class aircraft. Icing intensity is based on cloud parameters such as the liquid water concentration (LWC), temperatures, and drop size distribution, as well as aircraft type, speed, and duration within the clouds.

Observations of the tops of layered stratiform clouds can provide clues for the presence of possible icing conditions. Aside from simply observing sub-freezing CTTs, the presence of super-cooled liquid water (SLW) can be observed in multi-spectral satellite imagery due to its accumulation near the cloud tops (Rauber and Tokay 1991). Early attempts to identify SLW regions showed promise (Ellrod 1996; Thompson et al. 1997). Some shortcomings of these early icing products, however, were the obscuration of SLW clouds by cirrus, as well as the lack of information regarding icing intensity and the base of the icing layer.

The GOES icing product was later enhanced by an improved four-band icing algorithm and the addition of the Cloud Top Pressure (CTP) derived from the GOES sounder to show the maximum altitude ranges of SLW clouds (Ellrod and Bailey 2007). The improved four-band icing algorithm (referred to as “ICECAP”) corrected the visible imagery brightness for viewing angle and filtered much of the false icing due to cirrus day and night using empirical thresholds based on the visible and 3.9-, 10.7- and 12.0-µm IR data. There was still some false detection present, such as when there was thin cirrus over a snow cover. Verification for a 2-year period ending April 2005 showed excellent POD, especially at altitudes < 12 kft, but PODn (the ability to correctly observe non-icing regions) was unacceptably low.

A system that derived pixel level cloud properties based on four spectral GOES bands at 0.63, 3.9, 10.8, and 12.0 µm was used to successfully identify SLW clouds (Smith et al. 2000). If a cloud was determined to be liquid with temperatures below 0 °C, it was denoted as SLW with the potential to produce aircraft icing. The algorithm also worked well in the presence of thin cirrus (coverage < 5%), detecting 95% of SLW cases based on PIREPs. Later improvements to the algorithm (Smith et al. 2003; Minnis 2004) led to the prototype candidate for the GOES-R program, described by Smith et al. (2012). The system produces a Flight Icing Threat (FIT) index based on a combination of a satellite-derived icing mask, icing probability and intensity. An example of the GOES FIT product is shown by Fig. 12. Probabilities (low, medium, and high) are shown for light (LGT) and moderate or greater (MOG) icing.
Fig. 12

Flight icing threat (FIT) derived from GOES at 1745 UTC, 8 November 2008

(From Smith et al. 2012)

6 Concluding Remarks

Since the 1970s, data from geostationary meteorological satellites have proven their worth as valuable supplements to land-based and aircraft observations in the detection and short-range prediction of aviation hazards such as fog and low stratus, turbulence, thunderstorms, volcanic ash, and aircraft icing. Recent improvements in spectral and spatial resolution, image frequency, and navigational accuracy and stability have increased their value even more. For example, the new imagers can provide full disk coverage every 10–15 min. Not to be overlooked is the recent ability to relay weather products directly to aircraft cockpits for display using electronic flight bags (EFBs). Initially, only text products could be transmitted but rapid advances in ground-to-air satellite communications now allow raw imagery and derived products to be displayed in real time, resulting in increased situational awareness by aircrews. One example is the CDO product described in Sect. 4.2.1, which is now commercially available for satellite uplink to aircraft (Kessinger et al. 2017).

In addition to multi-spectral imagers, geostationary sounders such as GOES-VAS have shown the potential to monitor destabilization to predict possible thunderstorm development or microburst activity and to provide gradient and atmospheric motion vectors (AMVs) in remote areas for use in NWP. With the exception of China’s Geostationary Interferometric Infrared Sounder (GIIRS) on the FY-4 series (Yang et al. 2017), there are currently no sounders in geostationary orbit, but hyperspectral sounders (with many hundreds of spectral bands) are planned for future spacecraft such as Europe’s MTG, and possibly future GOES. In conclusion, the future looks bright for monitoring inflight and surface aviation weather conditions from space.

NOTE: Details concerning the suite of derived products from GOES-16 ABI and GLM are included in the Algorithm Theoretical Bases Documents (ATBD) available from NOAA/NESDIS at: (1) https://www.goes-r.gov/products/baseline.html (for Baseline products such as cloud top height, and volcanic ash detection), and: (2) https://www.goes-r.gov/products/option2.html (for future products such as aircraft icing, fog and low clouds, SO2 detection, convective initiation, enhanced-V/overshooting tops, and tropopause folding turbulence prediction).

Notes

Acknowledgements

The authors would like to thank Dr. Michael Pavolonis (NOAA/CIMSS) for his contributions to the section on volcanic ash detection. Many of the images in this paper were obtained from the University of Wisconsin CIMSS blog pages on the use of improved satellite image data from GOES-R (16) and Himiwari. We also acknowledge the comments of two anonymous reviewers that greatly improved the quality of the paper.

References

  1. Ackerman, S., Schreiner, A., Schmit, T., Woolf, H., Li, J., & Pavolonis, M. (2008). Using the GOES Sounder to monitor upper level SO2 from volcanic eruptions. Journal of Geophysical Research: Atmospheres, 113.  https://doi.org/10.1029/2007jd009622.
  2. Adler, R. F., & Fenn, D. D. (1979). Thunderstorm intensity as determined from satellite data. Journal of Applied Meteorology, 18, 502–517.CrossRefGoogle Scholar
  3. Atkins, N. T., & Wakimoto, R. M. (1991). Wet microburst activity over the southeastern United States: Implications for forecasting. Weather and Forecasting, 6, 470–482.CrossRefGoogle Scholar
  4. Bachmeier, S. (2017a). Tornadoes and Large Hail in Minnesota and Wisconsin. CIMSS Satellite Blog, Cooperative Institute for Meteorological Satellite Studies, 16 May 2017. http://cimss.ssec.wisc.edu/goes/blog/archives/category/severe-convection/page/6.
  5. Bachmeier, S. (2017b). Eruption of Kambalny Volcano in Kamchatka, Russia. CIMSS Satellite Blog, Cooperative Institute for Meteorological Satellite Studies, 25 March 2017. http://cimss.ssec.wisc.edu/goes/blog/archives/category/volcanic-activity/page/2.
  6. Bass, R. (2017). The FAA’s convective weather research program. In AMS 18th Conf. on Aviation, Range and Aerospace Meteorology (ARAM), 23 January 2017, Seattle, Washington. https://ams.confex.com/ams/97Annual/webprogram/Paper313429.html.
  7. Bedka, K., Brunner, J., Dworak, R., Feltz, W., Otkin, J., & Greenwals, T. (2010). Objective satellite-based detection of overshooting tops using infrared window channel brightness temperature gradients. Journal of Applied Meteorology and Climatology, 49, 181–202.CrossRefGoogle Scholar
  8. Bessho, K., et al. (2016). An introduction to Himiwari-8/9—Japan’s new-generation geostationary meteorological satellites. Journal of the Meteorological Society of Japan, 94, 151–183.CrossRefGoogle Scholar
  9. Bézy, J.-L., Aminou, D., Bensi, P., Stuhlman, R., Tjemkes, S., & Rodriguez, A. (2005). Meteosat Third Generation. European Space Agency Bulletin, 123, 28–32.Google Scholar
  10. Byers, H. R., & Braham, R. R. (1949). The thunderstorm (p. 247). Washington, DC: U. S. Government Printing Office.Google Scholar
  11. Calvert, C., & Pavolonis, M. (2011). GOES-R Advanced Baseline Imager (ABI) Theoretical Basis Document for Low Cloud and Fog, Version 2.0. NOAA/NESDIS Center for Satellite Applications and Research, p. 72.Google Scholar
  12. Caracena, F., & Flueck, J. A. (1988). Classifying and forecasting microburst activity in the Denver area. Journal of Aircraft, 25, 525–530.CrossRefGoogle Scholar
  13. Casadevall, T. (1994). The 1989–1990 eruption of Redoubt Volcano, Alaska: Impacts on aircraft operations. Journal of Volcanology and Geothermal Research, 62, 301–316.CrossRefGoogle Scholar
  14. Chai, T., Crawford, A., Stunder, B., Pavolonis, M., Draxler, R., & Stein, A. (2017). Improving volcanic ash predictions with the HYSPLIT dispersions model by assimilating MODIS satellite retrievals. Atmospheric Chemistry and Physics, 17, 2865–2879.CrossRefGoogle Scholar
  15. Clarisse, L., Prata, F., Lacour, J.-L., Hurtmans, D., Clerbaux, C., & Coheur, P.-F. (2010). A correlation method for volcanic ash detection using hyperspectral infrared measurements. Geophysical Research Letters, 37, L198065. (p. 5).CrossRefGoogle Scholar
  16. Conover, J. H. (1964). The identification and significance of orographically induced clouds observed by TIROS satellites. Journal of Applied Meteorology, 3, 226–234.CrossRefGoogle Scholar
  17. DeMaria, M., DeMaria, R. T., Knaff, J., & Molenar, D. (2012). Tropical cyclone lightning and rapid intensity change. Monthly Weather Review, 140, 1828–1842.CrossRefGoogle Scholar
  18. Donovan, M., Williams, E., Kessinger, C., Blackburn, G., Herzegh, P., Bankert, R., et al. (2008). The identification and verification of hazardous convective cells over oceans using visible and infrared satellite observations. Journal of Applied Meteorology and Climatology, 47, 164–184.CrossRefGoogle Scholar
  19. Dworak, R., Bedka, K., Brunner, J., & Feltz, W. (2012). Comparison between GOES-12 overshooting-top detections, WSR-88D radar reflectivity and severe storm reports. Weather and Forecasting, 27, 684–699.CrossRefGoogle Scholar
  20. Eckert, M. (2017). GOES-16 enhancements to IDSS for San Francisco International Airport (SFO) on 03/03/17. NWA Specialized Operations CommitteeIDSS Success Stories. http://nwas.org/wp-content/uploads/2016/08/DSS-Success-Story-SFO_GOES-16-1.pdf.
  21. Ellrod, G. P. (1985). Indicators of high altitude, non-convective turbulence observed in satellite images. In Proc. of the 2nd Intl. Conf. on the Aviation Weather System (pp. 277–284). Boston, MA: Amer. Meteor. Soc.Google Scholar
  22. Ellrod, G. P. (1987). Identifying high altitude mountain wave turbulence and strong chinook wind events with satellite imagery. In Proc. Of the AIAA 25th Aerospace Sciences Meeting, Jan. 12–15, Reno (p. 7). Washington, DC: Amer. Inst. For Aeronautics and Astronautics.Google Scholar
  23. Ellrod, G. P. (1989). Environmental conditions associated with the Dallas microburst storm determined from satellite soundings. Weather and Forecasting, 4, 469–484.CrossRefGoogle Scholar
  24. Ellrod, G. P. (1990). A water vapor feature related to severe thunderstorms. National Weather Digest, 15, 19–29.Google Scholar
  25. Ellrod, G. P. (1995). Advances in the detection of fog at night using GOES multispectral infrared imagery. Weather and Forecasting, 10, 606–619.CrossRefGoogle Scholar
  26. Ellrod, G. P. (1996). The use of GOES-8 multi-spectral imagery for the detection of aircraft icing regions. In 8th Conf. on Satellite Meteorology and Oceanography, Atlanta, GA, Amer. Meteor. Soc., (pp. 168–171).Google Scholar
  27. Ellrod, G. P. (2002). Estimation of low cloud base heights at night from satellite infrared and surface temperature data. National Weather Digest, 26, 39–44.Google Scholar
  28. Ellrod, G. P. (2005). Remote sensing of volcanic ash. Tutorial developed for National Weather Association Remote Sensing Committee. http://www.ellrodweather.com/volcano/ash.htm.
  29. Ellrod, G. P., & Bailey, A. (2007). Assessment of aircraft icing potential and maximum icing altitude from geostationary meteorological satellite data. Weather and Forecasting, 22, 160–174.CrossRefGoogle Scholar
  30. Ellrod, G. P., Connell, B., & Hillger, D. (2003). Improved detection of airborne volcanic ash using multi-spectral infrared satellite data. Journal of Geophysical Research, 108(D12), 4356.CrossRefGoogle Scholar
  31. Ellrod, G. P., & Gultepe, I. (2007). Inferring low cloud base heights at night for aviation using satellite infrared and surface temperature data. Pure and Applied Geophysics, 164, 1193–1205.CrossRefGoogle Scholar
  32. Ellrod, G. P., Maturi, E., & Steger, J. (1989). Detection of fog at night using dual channel GOES-VAS imagery. In Proc. 12th Conf. on Wea. Analysis and Forecasting, Monterey, California, Amer. Meteor. Soc. (pp. 515–520).Google Scholar
  33. Ellrod, G. P., & Schreiner, A. (2004). A first look at volcanic ash detection in the GOES-12 era. In 11th Conf. on Aviation, Range, and Aerospace Meteorology, Hyannis, MA, 4–8 October 2004, Paper #8.13. Boston, MA: Amer. Meteor. Soc.Google Scholar
  34. Eyre, J. R., Brownscombe, J. L., & Allam, R. J. (1984). Detection of fog at night using Advanced Very High Resolution Radiometer (AVHRR) imagery. Meteorological Magazine, 114, 187–201.Google Scholar
  35. Federal Aviation Administration (FAA). (2017). NextGen Weather FAQ: Weather Delays. https://www.faa.gov/nextgen/programs/weather/faq/.
  36. Fritz, S. (1965). The significance of mountain lee waves as seen from satellite pictures. Journal of Applied Meteorology, 4, 31–37.CrossRefGoogle Scholar
  37. Goodman, S. J., Blakeslee, R. J., Koshak, W. J., Mach, D., Bailey, J., Buechler, D., et al. (2013). The GOES-R Geostationary Lightning Mapper (GLM). Atmospheric Research, 125–126, 34–49.CrossRefGoogle Scholar
  38. Gravelle, C., Mecikalski, J., Line, W., Bedka, K., Petersen, R., Sieglaff, J., et al. (2016). Satellite convective toolkit to “bridge the gap” between severe weather watches and warnings: An example from the 20 May 2013 Moore, Oklahome tornado outbreak. Bulletin of the American Meteorological Society, 97(1), 69–84.CrossRefGoogle Scholar
  39. Gultepe, I., Pagowski, M., & Reid, J. (2007). Using surface data to validate a satellite based fog detection scheme. Weather and Forecasting, 22, 444–456.CrossRefGoogle Scholar
  40. Gurka, J. J. (1978). The use of enhanced visible imagery for predicting the time of fog dissipation. In Proc. Conf. on Weather Forecasting and Analysis and Aviation Meteorology, Silver Spring, MD, Amer. Meteor. Soc., (pp. 343–346).Google Scholar
  41. Han, H., Lee, S., Im, J., Kim, M., Lee, M.-I., Ahn, M., et al. (2015). Detection of convective initiation using Meteorological Imager onboard Communications, Ocean, and Meteorological Satellite based on machine learning approaches. Remote Sensing, 7, 9184–9204.CrossRefGoogle Scholar
  42. Hartung, D. C., Sieglaff, J. M., Cronce, L. M., & Feltz, W. F. (2013). An intercomparison of UW cloud-top cooling rates with WSR-88D radar data. Weather and Forecasting, 28, 463–480.CrossRefGoogle Scholar
  43. Heymsfield, G. M., & Blackmer, R. H. (1988). Satellite-observed characteristics of Midwest severe thunderstorm anvils. Monthly Weather Review, 116, 2200–2224.CrossRefGoogle Scholar
  44. Hillger, D., Kopp, T., Lee, T., Lindsey, D., Seaman, C., Miller, S., et al. (2013). First-light imagery from Suomi NPP VIIRS. Bulletin of the American Meteorological Society, 94, 1019–1029.CrossRefGoogle Scholar
  45. Hufford, G., Salinas, L., Simpson, J., Barske, E., & Pieri, D. (2000). Operational implications of airborne volcanic ash. Bulletin of the American Meteorological Society, 81(4), 745–755.CrossRefGoogle Scholar
  46. Kessinger, C., Megenhardt, D., Blackburn, G., Olivo, J., Lin, L., Hoang, V., Nayote, M., Sievers, K., Ritter, A., Wolf, D., Matz, O., Scheinhartz, R., & Cahall, J. (2017). Displaying convective weather products on an electronic flight bag. Journal of Air Traffic Control, Fall Issue, 52–61.Google Scholar
  47. Kim, D., & Ahn, M. (2014). Introduction of the in-orbit test and its performance for the first meteorological imager of the Communications, Ocean and Meteorological Satellite. Atmospheric Measurement Techniques, 7, 2471–2485.CrossRefGoogle Scholar
  48. Knox, J., Bachmeier, S., Carter, W., Tarantino, J., Paulik, L., Wilson, E., Bechdol, G., & Mays, M. (2010). Transverse cirrus bands in weather systems: a grand tour of an enduring enigma. Weather Magazine, 65(2), 35–41.  https://doi.org/10.1002/wea.417.CrossRefGoogle Scholar
  49. Lay, E. H., Holzworth, R. H., Rodger, C. J., Thomas, J. N., Pinto, O., & Dowden, R. L. (2004). WWLL global lightning detection system: Regional validation study in Brazil. Geophysical Research Letters, 31, L03102.  https://doi.org/10.1029/2003GL018882.Google Scholar
  50. Lee, Y.-K., Li, Z., Li, J., & Schmit, T. (2014). Evaluation of the GOES-R ABI LAP retrieval algorithm using the GOES-13 sounder. Journal of Atmospheric and Oceanic Technology, 31(1), 3–19.CrossRefGoogle Scholar
  51. Lee, T. F., Turk, F. J., & Richardson, K. (1997). Stratus and fog products using GOES-8-9 3.9 µm data. Weather and Forecasting, 12, 664–677.CrossRefGoogle Scholar
  52. Li, Z., Li, J., Menzel, W. P., Schmit, T. J., Nelson, J. P., III, Daniels, J., et al. (2008). GOES sounding improvement and applications to severe storm nowcasting. Geophysical Research Letters, 35, L03806.  https://doi.org/10.1029/2007gl032797.Google Scholar
  53. Lilly, D. K. (1978). A severe downslope windstorm and aircraft turbulence event induced by a mountain wave. Journal of the Atmospheric Sciences, 34, 59–77.CrossRefGoogle Scholar
  54. Lindstrom, S. (2017). Harvey and GLM Lightning. CIMSS Satellite Blog, Cooperative Institute for Meteorological Satellite Studies, 25 August 2017. http://cimss.ssec.wisc.edu/goes/blog/archives/date/2017/08/page/2.
  55. Liu, C., & Heckman, S. (2011). The application of total lightning detection and cell tracking for severe weather prediction. In 91st American Meteorological Society Annual Meeting, Seattle, 1–10.Google Scholar
  56. Martin, D. W., Kohrs, R. A., Mosher, F. R., Medaglia, C. M., & Adamo, C. (2008). Over-ocean validation of the Global Convective Diagnostic. Journal of Applied Meteorology and Climatology, 47, 525–543.CrossRefGoogle Scholar
  57. McCann, D. W. (1983). The enhanced-V, a satellite observable severe storm signature. Monthly Weather Review, 111, 887–894.CrossRefGoogle Scholar
  58. Mecikalski, J., & Bedka, K. (2006). Forecasting convective initiation by monitoring the evolution of moving cumulus in daytime GOES imagery. Monthly Weather Review, 134, 49–77.CrossRefGoogle Scholar
  59. Mecikalski, J., Williams, J. K., Jewett, C. P., Ahijevych, D., LeRoy, A., & Walker, J. R. (2015). Probabilistic 0–1-h convective initiation nowcasts that combine geostationary satellite observations and numerical weather prediction model data. Journal of Applied Meteorology and Climatology, 54, 1039–1059.  https://doi.org/10.1175/JAMC-D-14-0129.1.CrossRefGoogle Scholar
  60. Menzel, W. P., Holt, F. C., Schmit, T. J., Aune, R., Schreiner, A. J., Wade, G. S., et al. (1998). Application of GOES-8/9 soundings to weather forecasting and nowcasting. Bulletin of the American Meteorological Society, 10, 2059–2077.CrossRefGoogle Scholar
  61. Menzel, W. P., & Purdom, J. F. W. (1994). Introducing GOES-I: The first of a new generation of geostationary operational environmental satellites. Bulletin of the American Meteorological Society, 75, 757–780.CrossRefGoogle Scholar
  62. Merk, D., & Zinner, T. (2013). Detection of convective initiation using Meteosat SEVIRI implementation in and verification with the tracking and nowcasting algorithm Cb-TRAM. Atmospheric Measurement Techniques, 6, 1903–1918.CrossRefGoogle Scholar
  63. Miller, T., & Casadevall, T. (2000). Volcanic ash hazards to aviation: Encyclopedia of volcanoes. In H. Sigurdsson (Ed.), (pp. 915–930) San Diego, California: Academic.Google Scholar
  64. Miller, S. D., Schmit, T., Seaman, C., Lindsey, D., Gunshor, M., Kohrs, R., et al. (2016). A sight for sore eyes: The return of true color to geostationary satellites. Bulletin of the American Meteorological Society, 97, 1803–1816.CrossRefGoogle Scholar
  65. Minnis, P., et al. (2004). Real-time cloud, radiation, and aircraft icing parameters from GOES over the USA. In 13th Conf. on Satellite Oceanography and Meteorolology, Norfolk, Virginia, Amer. Meteor. Soc., P7.1. http://ams.confex.com/ams/pdfpapers/79179.pdf.
  66. Smith, W. L. Jr, Minnis, P., Bernstein, B. C., McDonough, F., & Khaiyer, M. M. (2003). Comparison of super-cooled liquid water cloud properties derived from satellite and aircraft measurements. In Proc. in-flight icing/de-icing int. conf., Chicago, IL, Federal Aviation Administration 2003-01-2156.Google Scholar
  67. Smith, W. L. Jr, Minnis, P., & Young, D. F. (2000). An icing product derived from operational satellite data. In Proc. AMS 9th conference on aviation, range and aerospace meteorology, 11–15 September 2000 (pp. 256–259). Orlando, FL: Amer. Meteor. Soc.Google Scholar
  68. Mohr, T. (2014). Preparing for the use of new generation geostationary meteorological satellites. WMO Bulletin, 63, 42–44.Google Scholar
  69. Morel, P., Desbois, M., & Szewach, G. (1978). A new insight into the troposphere with the water vapor channel of Meteosat. Bulletin of the American Meteorological Society, 59, 711–714.CrossRefGoogle Scholar
  70. Mosher, F. R. (2001). A satellite diagnostic of global convection. In 11th Conf. on Satellite Meteorology and Oceanography (pp. 416–419). Madison, Wisconsin: Amer. Meteor. Soc.Google Scholar
  71. Mosher, F. R. (2002). Detection of deep convection around the globe. In 10th Conf. on Aviation, Range, and Aerospace Meteorology (pp. 289–292). Portland, Oregon: Amer. Meteor. Soc.Google Scholar
  72. Negri, A. J., & Adler, R. F. (1981). Relation of satellite-based thunderstorm intensity to radar-estimated rainfall. Journal of Applied Meteorology, 20, 288–300.CrossRefGoogle Scholar
  73. Pavolonis, M. J. (2010). Advances in extracting cloud composition information from spaceborne infrared radiances—A robust alternative to brightness temperatures. Part I: Theory. Journal of Applied Meteorology and Climatology, 49(9), 1992–2012.CrossRefGoogle Scholar
  74. Pavolonis, M. J. (2018). Personal communication.Google Scholar
  75. Pavolonis, M. J., Feltz, W. F., Heidinger, A. K., & Gallina, G. M. (2005). A daytime complement to the reverse absorption technique for improved automated detection of volcanic ash. Journal of Atmospheric and Oceanic Technology, 23, 1422–1444.CrossRefGoogle Scholar
  76. Pavolonis, M. J., Heidinger, A., & Sieglaff, J. (2013). Automated retrievals of volcanic ash and dust cloud properties from upwelling infrared measurements. Journal of Geophysical Research-Atmospheres, 118(3), 1436–1458.CrossRefGoogle Scholar
  77. Pavolonis, M. J., Sieglaff, J. M., & Cintineo, J. L. (2015a). Spectrally enhanced cloud objects (SECO): A generalized framework for automated detection of volcanic ash and dust clouds using passive satellite measurements, Part I: Multispectral analysis. Journal of Geophysical Research: Atmospheres, 120, 7813–7841.Google Scholar
  78. Pavolonis, M. J., Sieglaff, J. M., & Cintineo, J. L. (2015b). Spectrally enhanced cloud objects (SECO): A generalized framework for automated detection of volcanic ash and dust clouds using passive satellite measurements, part II: Cloud object analysis and global application. Journal of Geophysical Research: Atmospheres, 120, 7842–7870.Google Scholar
  79. Prata, F. (1989). Observations of volcanic ash clouds in the 10–12 µm window using AVHRR/2 data. International Journal of Remote Sensing, 10, 751–761.CrossRefGoogle Scholar
  80. Prata, F., Schreiner, A., Schmit, T. J., & Ellrod, G. P. (2004). First measurements of volcanic sulfur dioxide from the GOES Sounder: Implications for improved aviation safety. In Proceedings, 2nd Intl. Conf. on Volcanic Ash and Aviation Safety, 21–24 June 2004, Alexandria, Virginia, Paper number 3.7.Google Scholar
  81. Pryor K. (2010). Recent developments in microburst nowcasting using GOES. In 17th conf. on satellite meteorology and oceanography (p. 9.7). Annapolis, Maryland: Amer. Meteor. Soc.Google Scholar
  82. Pryor K. (2012). Microburst nowcasting applications of GOES. In 18th conf. on satellite meteorology, oceanography and climatology (p. 471). New Orleans, Louisiana: Amer. Meteor. Soc.Google Scholar
  83. Pryor, K. (2014). Downburst prediction applications of meteorological geostationary satellites. In Proc. SPIE conf. on remote sensing of the atmosphere, clouds, and precipitation V, Beijing, China.  https://doi.org/10.1117/12.2069283.
  84. Pryor, K. (2015). Progress and developments of downburst prediction applications of GOES. Weather and Forecasting, 30, 1182–1200.CrossRefGoogle Scholar
  85. Pryor, K. (2017). Advances in downburst monitoring and prediction with GOES-16. In 17th conf. on mesoscale processes, San Diego, CA, Amer. Meteor. Soc., Paper No. 10.6.Google Scholar
  86. Pryor, K., & Ellrod, G. P. (2004a). Recent improvements to the GOES microburst products. Weather and Forecasting, 19, 582–594.CrossRefGoogle Scholar
  87. Pryor, K., & Ellrod, G. P. (2004b). WMSI—A new index for forecasting wet microburst severity. Electronic Journal Of Operational Meteorology, 5(3), 1–25.Google Scholar
  88. Purdom, J. F. W. (1973). Meso-highs and satellite imagery. Monthly Weather Review, 101, 180–181.CrossRefGoogle Scholar
  89. Purdom, J. F. W. (1976). Some uses of high-resolution GOES imagery in the mesoscale forecasting of convection and its behavior. Monthly Weather Review, 104, 1474–1483.CrossRefGoogle Scholar
  90. Rauber, R., & Tokay, A. (1991). An explanation for the existence of supercooled water at the top of cold clouds. Journal of Atmospheric Science, 48, 1005–1023.CrossRefGoogle Scholar
  91. Rose, W. I., Kostinski, A. B., & Kelley, L. (1995). Real time C band radar observations of 1992 eruption clouds from Crater Peak/Spurr Volcano, Alaska. U. S. Geological Survey Bulletin 2139, p. 19.Google Scholar
  92. Rose, W. I., & Mayberry, G. C. (2000). Use of GOES thermal infrared imagery for eruption scale measurements, Soufriere Hills, Montserrat. Geophysical Research Letters, 27, 3097–3100.CrossRefGoogle Scholar
  93. Schmetz, J., Pili, P., Tjemkes, S., Just, D., Kerkmann, J., Rota, S., et al. (2002). An introduction to Meteosat Second Generation (MSG). Bulletin of the American Meteorological Society, 83, 977–992.CrossRefGoogle Scholar
  94. Schmetz, J., Tjemkes, S. A., Gube, M., & van de Berg, L. (1997). Monitoring deep convection and convective overshooting with METEOSAT. Advances in Space Research, 19, 433–441.CrossRefGoogle Scholar
  95. Schmit, T., Goodman, S., Gunshor, M., Sieglaff, J., Heidinger, A., Bachmeier, S., et al. (2015). Rapid refresh information of significant events: Preparing users for the next generation of geostationary operational satellites. Bulletin of the American Meteorological Society, 96, 561–575.CrossRefGoogle Scholar
  96. Schmit, T., Griffith, P., Gunshor, M., Daniels, J., Goodman, S., & Lebair, W. (2017). A closer look at the ABI on the GOES-R series. Bulletin of the American Meteorological Society, 98(4), 681–698.CrossRefGoogle Scholar
  97. Schmit, T., Gunshor, M., Menzel, W. P., Gurka, J., Li, J., & Bachmeier, S. (2005). Introducing the next-generation Advanced Baseline Imager on GOES-R. Bulletin of the American Meteorological Society, 86, 1079–1096.CrossRefGoogle Scholar
  98. Schmit, T., Li, J., Gurka, J. J., Goldberg, M. D., Schrab, K. J., Li, J., et al. (2008). The GOES-R advanced baseline imager and the continuation of current sounder products. Journal of Applied Meteorology and Climatology, 47, 2696–2711.CrossRefGoogle Scholar
  99. Schultz, C. J., Petersen, W. A., & Carey, L. D. (2009). Preliminary development and evaluation of lightning jump algorithms for the real-time detection of severe weather. Journal of Applied Meteorology and Climatology, 48, 2543–2563.CrossRefGoogle Scholar
  100. Seftor, C. J., Hsu, N., Herman, J., Bhartia, P., Torres, O., Rose, W., et al. (1997). Detection of volcanic ash clouds from Nimbus 7/Total Ozone Mapping Spectrometer. Journal of Geophysical Research, 102(D14), 16749–16760.CrossRefGoogle Scholar
  101. Sikdar, D. N., Suomi, V. E., & Anderson, C. E. (1970). Convective transport of mass and energy in severe storms over the United States—An estimate from a geostationary altitude. Tellus, 22, 521–532.CrossRefGoogle Scholar
  102. Smith, W. L., Jr., Minnis, P., Fleeger, C., Spangenberg, D., Palikonda, R., & Nguyen, L. (2012). Determining the Flight Icing Threat to aircraft with single-layer cloud parameters derived from operational satellite data. Journal of Applied Meteorology and Climatology, 51, 1794–1810.CrossRefGoogle Scholar
  103. Smith, W. L., Jr., Suomi, V., Menzel, W., Woolf, H., Sromovsky, L., Revercomb, H., et al. (1981). First sounding results from VAS-D. Bulletin of the American Meteorological Society, 62, 232–236.CrossRefGoogle Scholar
  104. Srivastava, R. C. (1987). A model of intense downdrafts driven by the melting and evaporation of precipitation. Journal of Atmospheric Science, 44, 1752–1773.CrossRefGoogle Scholar
  105. Suomi, V. E., & Krauss, R. J. (1978). The spin scan camera system: Geostationary meteorological satellite workhorse for a decade. Optical Engineering, 17, 6–13.CrossRefGoogle Scholar
  106. Tag, P., Bankert, R., & Brody, L. (2000). An AVHRR multiple cloud-type classification package. Journal of Applied Meteorology, 39, 125–134.CrossRefGoogle Scholar
  107. Thompson, G., Bullock, R., & Lee, T. F. (1997). Using satellite data to reduce spatial extent of diagnosed icing. Weather and Forecasting, 12, 185–190.CrossRefGoogle Scholar
  108. University of Wisconsin CIMSS. (2017). GOES-R Fog Product Examples, 17 October 2017. http://fusedfog.ssec.wisc.edu/?cat=18.
  109. Wakimoto, R. M. (1985). Forecasting dry microburst activity over the High Plains. Monthly Weather Review, 113, 1131–1143.CrossRefGoogle Scholar
  110. Walker, J. R., MacKenzie, W. M., Jr., Mecikalski, J. R., & Jewett, C. P. (2012). An enhanced geostationary satellite-based convective initiation algorithm for 0–2-h nowcasting with object tracking. Journal of Applied Meteorology and Climatology, 51, 1931–1949.  https://doi.org/10.1175/JAMC-D-11-0246.1.CrossRefGoogle Scholar
  111. Wimmers, A. J., & Feltz, W. (2010). Tropopause folding turbulence product. GOES-R Algorithm Theoretical Basis Document (ATBD). NOAA Center for Satellite Applications and Research. https://www.goes-r.gov/products/ATBDs/option2/Aviation_Turbulence_v1.0_no_color.pdf.
  112. Wimmers, A. J., Griffin, S., Gerth, J., Bachmeier, S., & Lindstrom, S. (2018). Observations of gravity waves with high-pass filtering in the new generation of geostationary imagers and their relation to aircraft turbulence. Weather and Forecasting, 33, 139–144.CrossRefGoogle Scholar
  113. Wimmers, A. J., & Moody, J. L. (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. Journal of Geophysical Research, 106(D15), 17115–17132.CrossRefGoogle Scholar
  114. Wimmers, A. J., & Moody, J. L. (2004a). Tropopause folding at satellite-observed spatial gradients: 1. Verification of an empirical relationship. Journal of Geophysical Research, 109, D19306.  https://doi.org/10.1029/2003JD004145.CrossRefGoogle Scholar
  115. Wimmers, A. J., & Moody, J. L. (2004b). Tropopause folding at satellite-observed spatial gradients: 2. Development of an empirical model. Journal of Geophysical Research, 109, D19307.  https://doi.org/10.1029/2003JD004146.CrossRefGoogle Scholar
  116. Yang, J., Zhang, Z., Wei, C., Lu, F., & Guo, Q. (2017). Introducing the new generation of Chinese geostationary weather satellites, Fengyun-4. Bulletin of the American Meteorological Society, 98, 1637–1659.CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.EWxC, LLCGranbyUSA
  2. 2.NOAA/NESDIS Center for Satellite Applications and ResearchCollege ParkUSA

Personalised recommendations