The Use of Northern Hemisphere- and Southern Hemisphere-Associated IR Enhancements to Assess Southern Hemispheric Tropical Cyclone Intensity: Tropical Cyclone Yasi
Bryce Touchstone
GR6753-501 Satellite and Radar Meteorology
Mississippi State University, Spring 2011
Introduction
In 1975, Vernon F. Dvorak created a technique, namely “The Dvorak Technique”, to assess current and future tropical cyclone intensity in the Atlantic Basin of the Northern Hemisphere. The technique consisted primarily of satellite imagery analysis by a meteorologist based upon cloud patterns, to depict both the current intensity of the tropical cyclone, as well as its projected future state(Dvorak, 1975). This method was based upon subjective methodology since the analysis was at the discretion of the human analyst.
Since the original Dvorak Technique relied on subjective analyst interpretation of both cloud patterns and empirical rules regarding cyclone intensity, and while this technique proved sufficient for operational meteorology, there were instances where discrepancies occurred between different analysis centres who were measuring the same tropical system. A more objective method was necessary to compensate in these instances. A computer-based algorithm was developed that analysed digital infrared imagery. This algorithm, named “The Objective Dvorak Technique”(ODT), has been modified since its creation to include certain of the original Dvorak rules, more constraints, and a component of time-averaging. This modified ODT is only applicable to tropical systems which have attained Tropical Storm or Hurricane strength(Velden, et. al., 1997).
While the algorithm component is beyond the scope of this project, there were IR Enhancement Curves associated with the ODT which I will use to analyse Tropical Cyclone Yasi during landfall of N Queensland, Australia, 2 February, 2011, with Category-5 Australian Tropical Cyclone Scale(ATCS) (Category-4 SSHS) strength. The use of the ODT has certain prerequisites, the afore-mentioned minimum strength category, and a well-defined eyewall; this is due to the fact that the ODT relies upon eyewall centre recognition to identify the centre of the storm. Once this is identified, the temperature (typically land or ocean surface) is coupled with temperature of surrounding bands of convective cloud-top temperatures at 4-km intervals from 24 km to 132 km from the eyewall centre to produce a CI (Current Intensity) number, corresponding to a T-Number, depicting both the current and projected future states of the tropical cyclone’s intensity (Velden, et. al., 1997). Since TC5- Yasi had a very well-defined eyewall, it has been decided to use IR Enhancements associated with the ODT to analyse the current intensity of TC5-Yasi.
After performing the IR Enhancements associated with the Northern Hemisphere method, an enhancement created by the Australian Bureau of Meteorology (BOM) will also be applied to compare the methods.
Climatological and Synoptic Setup
The 2010/2011 Southern Hemisphere Summer was marked by unprecedented flooding in Australia as a result of a very powerful La Niña sector of the ENSO. This La Niña peaked in early January, which was confirmed by Southern Oscillation Index (SOI) values, cloud patterns, and Pacific Ocean temperatures all showing a marked gradual return to ‘normal’ conditions subsequent mid-January, with further deterioration of La Niña expected to continue through the Southern Autumn (Australia BOM).
TC2-Anthony, a weaker tropical system forming a week prior to Yasi, caused widespread rainfall and flooding in Vanuatu, Solomon Islands, and New Caledonia, and struck with Category-2 (ATCS) force around Townsville, forcing its way inland and dumping 100mm of rain in 6 hours 1300km inland in the Outback. Yasi was identified by the Fiji Meteorological Service (FMS) as tropical disturbance “09F” 330km SW of Tuvalu on 26 January, 2011. Warm sea surface temperatures and relatively low wind shear were ahead of its projected path, and 27 January saw an upgrade to tropical depression status, with little to no intensification on 28 January. 30 January saw rapid intensification and upgrade to tropical storm status by the Joint Typhoon Warning Centre (JTWC), and shortly thereafter upgrade to Tropical Cyclone Yasi by the FMS. Tracking westward from 370km NE of Vanuatu, on 31 January, with ten minute sustained winds of at least 120km, Yasi attained Severe Tropical Cyclone Status(Wikipedia, BOM). Below is Yasi’s track.
Track of TC5-Yasi (Created by Keith Edkins and Iune using Wikipedia:WikiProject Tropical cyclones/Tracks. The background image is from NASA. Tracking data is from the Joint Typhoon Warning Center.)
Enhancement Process
Figure 1 – Mollweide IR Composite, 0900Z, .02.02.2011. (Greyscale)
Figure 2 – Mollweide IR Composite, 0900Z, .02.02.2011. (Inverse Greyscale)
Figures 1 and 2 are the unenhanced IR images of Tropical Cyclone Yasi (henceforth TC5-Yasi) during landfall on 2 Feb., ’11, S of Cairns, displayed in Greyscale and Inverse Greyscale, respectively. Inverse Greyscale will be retained for the subsequent enhancements. The first enhancement to be performed will be a simple ZA Enhancement to better show temperature contrast across the temperature spectrum of our interest. To obtain the temperature range necessary, I have applied contrasting colours at 10K-intervals beginning at 217K per ZA Enhancement table, Lecture 6, Slide 5 of my GR6753 lecture notes to identify the coldest cloud-top temperatures within the cyclone, and the resulting image and breakpoints with corresponding colours are shown in Figures 3 and 4, respectively. Note the well-defined eye in this image.
Figure 3 – Resulting image showing rough temperature estimates
Figure 4 – Breakpoints inserted at 10K-intervals and corresponding colours
( (167-177]K = Purple; (177-187]K = Blue; (187-197]K = Green; (197-207]K = Yellow; (207-217]K = Red)
Figure 5 - Temperature Histogram, 0900Z, .02.02.2011.
The temperature range can be seen in the Temperature Histogram for this particular image in Figure 5. This histogram is a rough diagram, but cannot be relied upon for absolute values as there may be ‘outliers’ which may not be visible to the eye as a result of the large y-axis scale, thus the systematic approach is necessary. As a result, the lowest (coldest) temperature interval is [197-187)K, thus a new range of 170K-315K will be selected, 315K because temperatures above 42C are not of interest to this project. The resultant image from this ZA Enhancement is shown in Figure 6.
Figure 6 – Image after ZA Enhancement applied.
The next step is the application of a BD Curve for Tropical Cyclone Intensity which is used in the ODT(Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin-Madison), Figure 7. An attempt to recreate this Enhancement Curve is displayed with breakpoint values along the colour table in Figure 8. The resultant image of the application of this BD Curve is shown in Figure 9. As you can see, given the poor resolution of the image, the application of this BD Curve is a disaster. Comparing our image to that of the CIMSS shows that higher resolution will produce much more useful data. While synoptic scale features are vaguely identifiable, the relatively narrower convective cloud band surrounding the eyewall cannot be discerned or evaluated using the BD Curve at this resolution.
Figure 7 – BD Curve, CIMSS
Figure 8 – Breakpoints inserted per BD Curve intervals
Figure 9 – BD Enhancement Curve
One other IR Enhancement to attempt is the Tropical Cyclone IR Enhancement Curve from the Australian Bureau of Meteorology. Below is a sample image of this curve and its corresponding colour/temperature values in Figures 10 and 11, respectively.
Figure 10 – Two Tropical Cyclones off of NT and WA, Australia.
"Satellite image processed by the Bureau of Meteorology from data received from the geostationary meteorological satellite MTSAT-1r operated by the Japan Meteorological Agency".
Figure 11 – BOM TC Key
The breakpoint/colour table and resultant image after the application of this IR Enhancement are shown in Figures 12 and 13, respectively.
Figure 12 – Breakpoint/Colour Table
Figure 13 – Satellite image after application of BOM TC IR Enhancement
It is obvious that the BOM TC IR Enhancement is much more successful at this resolution. The intense convective cloudbands to the E, N, and W can be seen much more clearly here. In the future, it would be much more useful to have higher resolution IR data. Perhaps with the launch of the new GOES-R, this will be available in the future.
References
-Australian Bureau of Meteorology (http://www.bom.gov.au)
-Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin-Madison (http://cimss.ssec.wisc.edu/tropic2/misc/other/faq/faq_enhance.html)
-“Tropical Cyclone Intensity Analysis and Forecasting from Satellite Imagery”, Vernon F. Dorak, 1975
-“Development of an Objective Scheme to Estimate Tropical Cyclone Intensity from Digital Geostationary Satellite Infrared Imagery”, Christopher S. Velden, Timothy L. Olander, Raymond M. Zehr
-Wikipedia (http://www.wikipedia.org)
.BT
