West Texas Downbursts: 13 August 2008

A downburst event over western Texas during the afternoon of 13 August 2008 served as another good example of the coordinated use of two predictive models: the Geostationary Operational Environmental Satellite (GOES)-sounder derived Microburst Windspeed Potential Index (MWPI) and GOES imager microburst algorithms, as compared to microburst risk algorithm output using 1 km MODIS data. During the afternoon of 13 August, a cluster of convective storms developed along the dryline over the Texas-New Mexico border. Several strong downbursts, produced by the convective storm cluster, occurred in close proximity to elevated GOES MWPI (Pryor 2008) and imager microburst risk (Pryor 2009) values. The imager microburst algorithm output, generated with MODIS data, indicated the strongest signal for downburst potential over western Texas. A downburst wind gust of 62 knots was recorded at Amherst West Texas Mesonet station while wind gusts of 43 to 48 knots were recorded by West Texas Mesonet stations in the Lubbock area. The ambient thermodynamic environment over western Texas was typical of the southern Great Plains with a deep, well-mixed boundary layer that favored evaporatively initiated downbursts. The MWPI algorithm incorporates relevant parameters for downburst potential, including convective available potential energy (CAPE) and boundary layer temperature and humidity gradients.

The images above are a GOES sounder-derived MWPI product and corresponding imager microburst risk product at 2000 UTC 13 August 2008. Apparent in the MWPI product image is a band of enhanced cumulus clouds and convective storm activity developing near the dryline over the Texas-New Mexico border that would evolve into a storm cluster over western Texas between 2100 and 2300 UTC. Based on a study by Weiss and Schultz (2006), a surface dewpoint difference of 1C was observed between Paducah, Texas and Dora, New Mexico (West Texas) mesonet stations at 2300 UTC, establishing the presence and location of the dryline. Associated with the convective storm cluster were downburst wind gusts of 38 to 62 knots (plotted in image) that were recorded by West Texas mesonet stations. Note that the downbursts occurred in close proximity to elevated MWPI values. The imager microburst risk product, with overlying radar reflectivity data from Lubbock, Texas NEXRAD (LBB) at 2215 UTC, displays a downburst-producing convective storm over Amherst, where a wind gust of 62 knots was observed. The convective storm cluster subsequently produced downburst wind gusts of 43 to 48 knots in the Lubbock area between 2325 and 2355 UTC. The Amherst downburst occurred in close proximity to high microburst risk (orange shading). High downburst risk was more apparent in the 2010 UTC MODIS image below as a darker orange shading immediately downstream of the convective storm over Amherst.

The GOES sounding above displayed an ambient thermodynamic environment typical of the southern Great Plains with a steep temperature lapse rate and a well-mixed boundary layer that favored the development of intense convective downdrafts as rain shafts descended into the subcloud layer. Comparison of the GOES sounder, imager and MODIS microburst products to the above sounding revealed that high risk values are associated with an "inverted V" profile (Johns and Doswell 1992). The dryline served as an initiating mechanism for convective storm activity and fostered a favorable environment for downbursts by enhancing vertical mixing (Ziegler and Hane 1993). Overall, the MWPI, imager, and MODIS microburst products effectively indicated the potential for strong convective wind gusts over western Texas during the afternoon of 13 August.

REFERENCES

Johns, R.H., and C.A. Doswell, 1992: Severe local storms forecasting. Mon. Wea. Rev., 121, 1134–1151.

Pryor, K.L., 2008: An initial assessment of the GOES Microburst Windspeed Potential Index. Preprints, 5th GOES Users' Conf., New Orleans, LA, Amer. Meteor. Soc.

Pryor, K.L., 2009: Microburst windspeed potential assessment: progress and developments. Preprints, 16th Conf. on Satellite Meteorology and Oceanography, Phoenix, AZ, Amer. Meteor. Soc.

Weiss, C.C., and D. M. Schultz, 2006: Synoptic and mesoscale influences on west Texas dryline development and associated convection. 23^rd Conf. on Severe Local Storms, St. Louis, MO, Amer. Meteor. Soc.

Ziegler, C.L., and C.E. Hane, 1993: An Observational Study of the Dryline. Mon. Wea. Rev., 121, 1134–1151.

Assessment of MODIS as a Data Source for Microburst Prediction

NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) is currently being evaluated as a data source for a multispectral algorithm that currently generates the GOES-11 Imager microburst product. The GOES-11 imager microburst algorithm, as described in Pryor (2009), is a predictive linear model (Caracena and Flueck 1988) based on parameter evaluation and pattern recognition techniques as employed in the severe convective storm forecasting process (Johns and Doswell 1992). At a spatial resolution of 1 km, the MODIS instrument is an improvement over the GOES-11 imager with a resolution of 4 km for the infrared bands. Thus, MODIS should provide better detail in the display of microburst risk. Two downburst events from 2008 will be compared to highlight the improvement in performance of MODIS over GOES-11 imager data. During the afternoon of 8 December 2008, clusters of convective storms developed over eastern New Mexico and western Texas, ahead of an upper-level cyclone. Strong downbursts, produced by the convective storm clusters, occurred in close proximity to elevated GOES-11 imager microburst risk values, as indicated by the GOES-11 product image at 2030 UTC. A comparison between the product images generated with GOES-11 and MODIS data revealed that the MODIS image indicated a stronger signal for microburst potential over western Texas during the afternoon of 8 December.


The images above are a MODIS microburst risk image (top) and corresponding GOES-11 microburst risk image at 2030 UTC 8 December 2008 (bottom) with overlying radar reflectivity data from Lubbock, Texas NEXRAD (LBB) at 2118 UTC. Apparent in the product images are clusters of convective storms over the western Texas Panhandle and over eastern New Mexico that would track eastward over western Texas during the following hour. Associated with the convective storm cluster near the New Mexico border were downburst wind gusts of 50 and 57 knots that were recorded by Plains and Denver City (West Texas) mesonet stations at 2115 UTC. Note that the downburst-producing convective storm near Denver City occurred in close proximity to elevated imager microburst risk values, displayed in the images as a progression from yellow to red shading indicating an increase from moderate to high risk. High downburst risk was more apparent in the MODIS image as a darker orange shading immediately downstream of the convective storm near Denver City. The downbursts were evaporatively initiated (Caracena and Flueck 1988) and resulted in the generation of a dust storm over western Texas that affected the Lubbock area (http://www.mesonet.ttu.edu/cases/DuststormI_120808/Dec-08-2008Duststorm.html).

Below is a comparison of the MODIS microburst risk image at 2035 UTC 9 August 2008 (top) and corresponding GOES-11 microburst risk image at 2130 UTC (bottom) with overlying radar reflectivity data from Phoenix, Arizona NEXRAD (IWA) at 2238 UTC. Apparent in the microburst product images is a convective storm over northern Maricopa County that produced a downburst (46 knot wind gust) at Horseshoe Lake Automated Local Evaluation in Real Time (ALERT) station at 2238 UTC. In a similar manner to the western Texas downbursts of 8 December, the strong downburst occurred in close proximity to a local maximum in risk, as indicated by the orange shading immediately downstream of the convective storm, especially apparent in the MODIS image with 1 km resolution.




Both cases demonstrate the strength of the MODIS-derived microburst risk image in the assessment of downburst potential, especially with evaporatively initiated downbursts. With 1 km spatial resolution, the MODIS image can serve as an example of the capability of future geostationary satellite sensors (i.e. GOES-R) in displaying multispectral derived images, especially images with utility in convective storm nowcasting.

REFERENCES

Caracena, F., and J.A. Flueck, 1988: Classifying and forecasting microburst activity in the Denver area. J. Aircraft, 25, 525-530.

Johns, R.H., and C.A. Doswell, 1992: Severe local storms forecasting. Mon. Wea. Rev., 121, 1134–1151.

Pryor, K.L., 2009: Microburst windspeed potential assessment: progress and developments. Preprints, 16th Conf. on Satellite Meteorology and Oceanography, Phoenix, AZ, Amer. Meteor. Soc.

Arizona Downbursts 2008

Summer of 2008 was rather convectively active as documented in a report released by the Flood Control District of Maricopa County (http://www.fcd.maricopa.gov/Rainfall/publications.aspx). Strong downbursts were observed by mesonet stations in the Arizona Automated Local Evaluation in Real Time (ALERT) domain between July and September 2008 and served as validation data for the Geostationary Operational Environmental Satellite (GOES)-11 imager microburst algorithm that employs brightness temperature differences (BTD) between band 3 (upper level water vapor, 6.7μm), band 4 (longwave infrared window, 10.7μm), and split window band 5 (12μm). A strong negative correlation has been found between 6.7μm brightness temperature (Tb) and layer-averaged relative humidity (RH) between the 200 and 500-mb levels. In the middle to upper troposphere, decreases in Tb are associated with increases in RH and, thus, a large BTD between bands 3 and 5 imply a large vertical relative humidity gradient. This difference in humidity between the mid-troposphere and the surface is a condition favorable for strong convective downdraft generation due to evaporational cooling as precipitation descends in the sub-cloud layer.

The first documented microburst in the ALERT domain occurred at Magma FRS in Pinal County during the evening of 19 July 2008 as shown in the GOES-11 imager derived microburst risk product at 0100 UTC 20 July 2008 (above) with overlying radar reflectivity from Phoenix, Arizona NEXRAD (IWA) at 0343 UTC. Apparent in the microburst product image is a convective storm over Pinal County, 45 miles southeast of Phoenix. The storm produced a downburst with a measured wind gust of 39 knots at Magma FRS at 0333 UTC and occurred in close proximity to a local maximum in risk (probability), as indicated by the orange shading immediately downstream of the convective storm.

During August 2008, a strong downburst was observed at Horseshoe Lake ALERT station. Apparent in the microburst product image (above) at 2130 UTC 9 August 2008 with overlying radar reflectivity from Phoenix, Arizona NEXRAD (IWA) at 2238 UTC is a convective storm over northern Maricopa County that produced a downburst at Horseshoe Lake at 2238 UTC. The strong downburst, with a wind gust of 46 knots, occurred in close proximity to a local maximum in risk, as indicated by the orange shading immediately downstream of the convective storm.

Finally, a strong downburst occurred at Gila Bend ALERT station during the afternoon of 11 September 2008. The above GOES-11 imager derived microburst risk product at 2000 UTC with overlying radar reflectivity from Phoenix, Arizona NEXRAD (IWA) at 0022 UTC 12 September displayed a convective storm over southwestern Maricopa County. The storm produced a downburst (42 knots) at Gila Bend at 0025 UTC and occurred in close proximity to a local maximum in risk, as indicated by the orange shading immediately downstream of the convective storm.

The ambient environments of all three downburst events were characterized by a well-mixed, convective boundary layer as exemplified by the "inverted-V" sounding profile above. This sounding, derived from Rapid Update Cycle (RUC) model analysis data at 0100 UTC 20 July 2008 over Magma FRS ALERT station, displays significant mid-level moisture overlying a dry surface layer that resulted from several hours of strong surface heating and resultant mixing. Apparent in the sounding profile is a large humidity gradient between the mid-troposphere and the surface that fostered strong convective downdraft generation due to evaporational cooling as precipitation descended in the sub-cloud layer.

Thus, based on preliminary analysis of three downburst events over central Arizona during the summer of 2008, the GOES-11 imager microburst product demonstrated effectiveness in the assessment of the potential for both dry and hybrid-type downbursts.

West Texas Windstorm: 8 February 2009

A convectively active late winter season over the Great Plains has proven fruitful for the assessment of the GOES-11 imager microburst risk product. During the evening of 8 February 2009, a line of convective storms tracked through eastern New Mexico and western Texas, producing several strong downbursts west of Lubbock. This event served as another good example of the utility of the GOES‐West (GOES‐11) imager microburst algorithm described in the previous blog entry (“Forecasting Convective Downburst Potential”, 29 January 2009). This downburst event occurred at the end of a day of boundary layer mixing due to a combination of strong surface heating and low-level wind shear, and thus, demonstrated the importance of the evolution of the convective mixed layer in downburst generation as reflected in the GOES microburst product imagery. The line of convective storms crossed the New Mexico border into Texas around 0100 UTC 9 February. The first downburst recorded in Texas was observed at Denver City West Texas Mesonet station with a wind gust of 47 knots at 0100 UTC, followed by a stronger downburst with a wind gust of 60 knots at 0125 UTC. Further downburst activity was observed at Anton (50 knots) and Reese Center (46 knots) mesonet stations west of Lubbock at 0225 UTC and 0235 UTC, respectively.



The image above is a recent example of the GOES‐11 imager microburst risk product at 0000 UTC 9 February 2009 with overlying radar reflectivity imagery from Lubbock (KLBB) NEXRAD at the time of downburst occurrence at Denver City, 0100 UTC. The product image was visualized by McIDAS-V software, available online at http://www.ssec.wisc.edu/mcidas/software/v/. The image was filtered to display only reflectivity higher than 35 dBZ to emphasize the heaviest precipitation cores where downbursts are likely to be generated. Apparent in the product image is the storm line crossing the Texas border, propagating into a region of high microburst probability as indicated by the progression from orange and red shading in the image. Clouds are represented by light to dark blue shaded areas in the product image. The next product image, valid at 0030 UTC with overlying radar reflectivity imagery at the time of downburst occurrence, shows the eastward progression of the storm line. 0221 UTC radar reflectivity data indicated a small bowing segment of the line northwest of Lubbock (L), associated with the downburst in progress at Anton.

Note that the Anton downburst again occurred in close proximity to elevated imager microburst risk values as displayed by darker orange shading.

This case demonstrated that the GOES-11 imager microburst algorithm output, when combined with radar reflectivity data into a composite product image, can effectively show forecasters where microbursts are likely. The GOES-11 microburst algorithm models the preconvective environment by utilizing brightness temperature differences between the midwave and longwave infrared channels to approximate favorable temperature and moisture gradients in the boundary layer that would enhance convective downdraft generation. More information about the GOES-11 microburst product can be found in the VISIT lesson titled “Forecasting Convective Downburst Potential Using GOES Sounder Derived Products”.