Arizona Downbursts: Recap and Look Ahead

As noted in a previous entry, "Arizona Downbursts 2008", analysis of three downburst events over central Arizona during the summer of 2008 revealed that the GOES-11 imager microburst product demonstrated effectiveness in the assessment of the potential for both dry and hybrid-type downbursts. In addition, it was also found these downburst events, observed by mesonet stations in the Arizona Automated Local Evaluation in Real Time (ALERT) domain between July and September 2008, served as effective validation data for the GOES-11 microburst product. This entry expands on the previous entry and looks ahead to the 2009 convective season validation effort over Arizona.


The images above are examples of southwestern U.S. GOES-11 imager microburst risk products that were generated prior to significant downburst events over central and southern Arizona during August and September 2008. Both images display convective storm activity developing over central and western Arizona during the afternoon that would produce downbursts during the following one to four hours. The product image at 2130 UTC 9 August 2008 (top) displays a favorable microburst environment near Horseshoe Lake ALERT station (AHL), where a strong convective wind gust of 46 knots would occur about one hour later at 2238 UTC. In a similar manner, the product image at 2000 UTC 11 September 2008 (bottom), indicates a moderate to high risk of microbursts near Gila Bend (AGL), where a downburst wind gust of 42 knots would be observed about four hours later at 0025 UTC 12 September.

Validation during the upcoming 2009 convective season will entail comparing surface observations of downburst wind gusts by Arizona ALERT stations to GOES-11 imager microburst and GOES-11 sounder Microburst Windspeed Potential Index (MWPI) product output generated one to three hours prior to each event. Wind gust magnitude for each downburst event will be compared directly to the closest representative microburst product output (BTD for imager product, MWPI for sounder product). In addition, radar reflectivity imagery will be utilized to verify that observed wind gusts are in fact associated with convective storms. Surface wind histograms from respective ALERT stations will be analyzed to identify the time and intensity of observed downbursts as demonstrated in the image below:

The above wind histogram at Horseshoe Lake ALERT station represents downburst occurrence as a sharp peak in wind gust speed near 1540 LST 9 August. This information, in conjunction with high radar reflectivity (>55 dBZ) associated with the parent convective storm, as displayed in the previous entry, confirmed that this wind event was clearly associated with a downburst. Thus, product validation will follow the methodology as outlined in Pryor (2009). It is expected that such a procedure should yield a statistically significant sample size, from which product performance should be effectively evaluated using classical statistical analysis. Validation involving three downburst events during the 2008 convective season did yield favorable results: a correlation of .50 that was statistically significant at the 84% confidence level. It is promising that such a small sample size indicated a significant correlation between GOES-11 microburst algorithm output and observed downburst wind gust speeds. More detail pertaining to this study can be reviewed in the following paper published in Arxiv.org: http://arxiv.org/abs/0904.0446.

Regional climatology has identified that the typical warm-season microburst environment over central Arizona can be best described as "hybrid" with an "inverted V" vertical profile characterized by significant CAPE, a mid-tropospheric moist layer, and a deep, dry convective boundary layer. These attributes are most effectively captured by both the GOES imager and sounder-derived MWPI products.

References

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

West Texas Windstorm: 14 August 2008

A convective high wind event over western Texas during the evening of 14 August 2008 served as another good example of the use of the Geostationary Operational Environmental Satellite (GOES)-West imager microburst algorithm. During the evening of 14 August, a cluster of convective storms that developed over eastern New Mexico merged with a larger mesoscale convective system (MCS) over the Texas Panhandle region, producing widespread high winds. A severe downburst, observed by Ralls (West Texas) Mesonet station, occurred in close proximity to elevated imager microburst risk values. The ambient thermodynamic environment over western Texas was typical of the southern Great Plains with a deep, well-mixed boundary layer that favored the development of intense convective downdrafts and resultant downburst generation.

The images above are the Geostationary Operational Environmental Satellite (GOES)-11 imager microburst risk product visualized with McIDAS-X (top) and McIDAS-V (bottom) software, valid at 2300 UTC 14 August 2008. The bottom image also displays radar reflectivity from Lubbock, Texas NEXRAD (KLBB) at 0235 UTC 15 August 2008, time of downburst occurrence. The imager microburst risk product (top) displays a large MCS over northwestern Oklahoma and the Texas Panhandle and a smaller convective storm cluster over eastern New Mexico. These two systems merged and produced widespread severe convective winds over western Texas. Also shown is the location of a severe downburst with an associated wind gust of 52 knots that occurred at Ralls (West Texas) Mesonet station, in close proximity to high microburst risk (orange shading). The bottom image, with overlying radar data, shows a bow echo over western Crosby county associated with the Ralls downburst. Note that the bow echo is located in a region of high microburst risk (orange shading), demonstrating the predictive value of this product for severe weather warning purposes.

The Rapid Update Cycle (RUC) sounding profile above, at Ralls, Texas at 2300 UTC 14 August, displayed a favorable thermodynamic environment with a steep temperature lapse rate and a well-mixed boundary layer. Comparison of the imager microburst product to the RUC sounding revealed that high risk values are associated with an "inverted V" profile. The West Texas Mesonet recorded 11 severe wind observations with this convective storm event, in which the majority of the high wind observations were associated with gust front passages. The downburst recorded at Ralls was distinctive in the meteogram above as a sharp peak in wind speed coincident with a peak in rainfall rate. Radar imagery at the time of downburst occurrence confirmed that the parent storm was overhead at the observing station. Thus, the Ralls downburst event was driven primarily by sub-cloud evaporation of precipitation and subsequent generation of negative buoyancy and downdraft acceleration- an "evaporatively initiated" downburst (Caracena and Flueck 1988).

REFERENCES

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

An Anomalous Nighttime Downburst Event

During the early morning of 10 March 2009, a strong downburst occurred over western Texas near Lubbock. The downburst, with an associated wind gust of 39 knots, was recorded at Anton West Texas Mesonet station, 20 miles northwest of Lubbock, at 0830 UTC. This event was unusual due the fact that the downburst originated from a cluster of showers with low reflectivity during the nighttime hours. A 100-mb (3000 feet) deep residual mixed layer (Stull 1988) was in place over the greater Lubbock area and fostered downdraft instability as precipitation descended in the sub-cloud layer. The GOES-11 imager microburst risk product effectively captured this favorable environment for downburst winds. This case may serve as an example of the effectiveness of the GOES imager product over the GOES-sounder Microburst Windspeed Potential Index (MWPI) in a region of widespread middle and high-level clouds.




The images above are a GOES sounder-derived MWPI product and corresponding imager microburst risk product at 0700 UTC 10 March 2009. Apparent in the GOES imager product is a cluster of showers over Hockley County, west of Lubbock. The location of downburst occurrence is marked with an "X". The closest clear-sky output brightness temperature differences (BTD) were indicated in excess of 40K (orange shading) over Lubbock. Previous validation (Pryor 2009) has established that output BTD greater than 40K was associated with wind gust potential of 40 knots or greater. However, since no sounding retrievals were available in the Lubbock area (REE) at 0700 UTC due to the presence of widespread mid-level clouds, no corresponding MWPI values were plotted. MWPI values of 11 to 14 (wind gust potential less than 35 knots) were plotted east of Plainview (PVS) and Ralls (RLS). Thus, the GOES imager product was able to indicate wind gust potential in the greater Lubbock area while sounder retrieval data over western Texas was very sparse due to the presence of middle and high clouds. In addition, the output BTD values, near 40K, more accurately indicated wind gust potential than was indicated by the MWPI product.

The above RUC model analysis sounding at 0700 UTC over Anton displayed the presence of the residual mixed layer that had developed and evolved during previous day. The layer, about 100 mb deep, provided sufficient downdraft instability that resulted from a steep, near dry-adiabatic lapse rate and large vertical relative humidity gradient. These conditions favored evaporative cooling and negative buoyancy generation as precipitation descended below the cloud base that was at a height near 10,000 feet. Thus, the cluster of relatively weak showers was capable of producing intense downdrafts due to a sub-cloud thermodynamic structure that provided very little inhibition. The residual mixed layer was overlying a conditionally unstable layer from the 800-mb level to the near the surface and a shallow stable boundary layer based at the surface. Surface observations at Anton (not shown) indicated a slight temperature increase at the time of downburst occurrence (0830 UTC), likely resulting from the mixing of warmer and drier air in the residual layer to the surface. Overall, the GOES-11 imager microburst risk product effectively indicated downburst potential in this nighttime, elevated mixed layer environment.

References

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

Stull, R.B., 1988: An introduction to boundary layer meteorology. Kluwer Academic Publishers, Boston, 649 pp.

Initial Assessment of the GOES Imager Microburst Product over Western Texas

Analysis of a few selected downburst cases over western Texas has provided an initial assessment of the performance of the GOES imager microburst product. Four cases between June and December 2008 featuring a total of eight downbursts in the West Texas Mesonet domain have been utilized. The results presented in this study will serve as a starting point for a much larger validation effort to be conducted during the summer of 2009.

Correlation was calculated by comparing output brightness temperature difference (BTD) values derived from GOES-11 and MODIS data to wind gust observations associated with downbursts. The cases selected included observations of severe winds (> 50 knots), in which downbursts occurred one to three hours after the generation of the product image. These cases provided the opportunity to assess the imager products as predictive tools for microburst potential. It was required in the product images that skies were clear over the location of downburst occurrence to preclude cloud contamination and unrepresentatively low risk values. A correlation of .59 was found between GOES-11 microburst risk values and wind gust magnitude. By a factor of two, this correlation was stronger than that calculated between MODIS output BTD and downburst wind gust measurements. Thus, over western Texas, GOES-11 imager data would be the optimal source to calculate and display microburst risk (probability).

The summer downburst events (6 of 8) were associated with high-reflectivity convective storms (>50 dBZ), highlighting the importance of precipitation loading in the downburst generation process (Johns and Doswell 1992). In most cases the downbursts occurred during the late afternoon and evening, near the end of a day of strong solar heating that resulted in the development of a deep convective boundary layer. The imager microburst products captured the evolution of these favorable environmental conditions.



The images above exemplify a typical summer downburst event over western Texas. The West Texas Mesonet meteogram at Turkey on 27 July 2008 displays the evolution of a deep convective mixed layer as an increase in dewpoint depression (DD) , in which the maximum surface DD (45F) occurred about one hour prior to the downburst. It is apparent in the meteogram that the development and evolution of the mixed layer is tied to strong insolation throught the day. The GOES-11 imager microburst product at 2030 UTC 27 July with overlying radar reflectivity imagery at 2307 UTC shows the downburst-producing convective storm near Turkey in a region of high risk. In general, severe downbursts, with wind gusts greater than 50 knots, are likely when output BTD values are greater than 45K.

REFERENCES

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