1. INTRODUCTION:

Storm Prediction Center Severe Weather (SVRGIS) Geographic Information System (GIS) applications are becoming increasingly popular in many organizations, including the federal government. Emergency operations have taken advantage of these capabilities. “GIS can be beneficial in many ways, but in the simplest of terms, it connects people to information through geography. Government can use GIS to store, manage, and access information about its facilities, people, and environment. It gives government officials a way to visualize data that helps them make decisions about project planning and economic development. It also allows them to disseminate a large quantity of information to the public in terms of where things and events are located." (PHELAN 2013) The Indian Meteorological Department (IMD)’s mission includes the protection of life and property through their National Weather Service’s (NWS’s) issuance of severe weather warnings. Over the past two spring and summer seasons, NOAA’s NWS Weather Forecast Office (WFO) in Glasgow, Montana has employed GIS information to significantly improve real-time verification.

This paper will discuss severe weather operations, the methodology employed in using GIS, during operations, compare verification from previous years to the present GIS-enhanced operations, and discuss specific case in October 2013 in which GIS was particularly helpful in severe weather operations. In addition to severe report data, other data such as geographic and archived meteorological data can be input into SVRGIS. The GIS user can select desired data by querying data by its attributes, location, or both. This is especially helpful when working with large datasets. One advantage of a GISbased system is the ability to view data in layers. This feature enables the user to quickly examine different variables (e.g., severe reports, municipal areas, transportation routes, and meteorological data) and examine any relationships between the variables.

This study’s purpose is to examine the methodology of importing severe report databases into a Geographical Information System (GIS) and highlight a few advantages and capabilities of using a GIS. Outlined are a few of the methods the author used to organize the severe report database into a GIS and include other data types (e.g., geographical) for comparison. A few severe report comparisons will be presented herein and will serve as examples to the many advantages and capabilities a GIS- based severe report database may have in analyzing severe data.

2. DATA COLLECTION/ DATABASE MANAGEMENT:

2a. Severe Reports

The severe reports used in this study originated from the National Climatic Data Center (NCDC) and were downloaded as text files (.txt) from the Storm Prediction Center (SPC) Tornado/Severe. It must be noted that description of the severe report attribute format used is available for download from the previously listed SPC web address. The reports are associated with descriptive data such as date, time, state(s) affected, Fscale, latitude and longitude coordinates, and monetary damage to name a few. In order to input the large number of recorded reports in a database table, the .txt files were imported into MS Office Access. Access converted the data into tables within an ArcGIS personal geodatabase. Necessary file-converting and file creating steps, not mentioned here, were completed in ArcGIS to create severe report files for display. A visual basic (.bas) script file was used to create and display tornado tracks from tornado touchdown and endpoints in ArcGIS.

Once the files with spatial coordinates were created, they were overlaid on projected maps of the contiguous United States. Other geographic data was input into the GIS for visual purposes. Density files were used, which were determined by values originating from a central grid cell with a radius distance turned 360 degrees for values of other grid cells. Lastly, point density layers and line density layers were created for different variables and some of their associated magnitudes (e.g., baseball-sized hail, significant tornado tracks) using the ArcGIS spatial analyst extension. It is worth noting that there were a few error values found in the severe database. Because of these errors, (e.g., null data values, missing or incomplete latitude/longitude coordinate pair), a few minor data filtering methods were required.

2b. Miscellaneous Data

Various types of geographic data can be input into SVRGIS and examined in many ways. Free data from the internet originated from different sources including the U.S. government and National Weather Service. Most of the miscellaneous data was geographic- based. A few example of data type include county and state maps, cities and towns, national weather Service county wamping Areas.

3. APPLICATION:

Meteorologists have investigated past severe weather report distribution (i.e. tornadoes, large hail, and severe convective wind reports) and have for many years qualitatively and quantitatively assessed areas where higher frequencies of severe weather have Occurred (PHELAN 2013). More sophisticated studies are now possible. For example, At the beginning of the second decade of October 2013 a very strong cyclone developed over the Bay of Bengal. Super cyclone 02B Phailin showed average wind speeds of up to 259 km/h making the storm category 5 cyclones, the highest category according to the Saffir-Simpson hurricane scale. Phailin became one of the strongest tropical cyclones ever recorded over the North Indian Ocean. Phailins track led towards the northeast of India, where the tropical cyclone in the federal state of Odisha made landfall and caused enormous damage.

PHAILIN – Super cyclonic Storm over Bay of Bengal

• Date: October 09-13, 2013 

• Maximum 1 min-sustained winds: 140 kt (259 kph) 

• Maximum wind gusts: 170 kt (315 kph) 

• Category 5 (according to Saffir-Simpson Hurricane

Scale) 

• Lowest pressure in storm center: 910 hPa on

October 10-11, 2013 

• Landfall: October 12, 2013, 15:45 UTC, near

Gopalpur (Odisha) 

• Storm surge: up to 3.5 m (according to media

reports) 

• Maximum significant wave height: > 50 ft (15 m) 

• Phailin was one of only 4 category 5 cyclones over

Bay of Bengal 

• Unexpected intensity (by most forecast models) 

• Rapid development within 24 hours from tropical storm into cat 4 cyclone 

Evolution of tropical cyclone Phailin

Phailin originated from a tropical disturbance that moved westward over the Andaman Sea. On October 9, 2013, the cloud complex formed a closed cyclonic circulation near the archipelago of the Andaman and Nicobar Islands and then intensified into a tropical storm. At 12 UTC the mean wind speed was 40 kt (74 kph) and on the Andaman Islands thunderstorms brought heavy rain already. At this time, many forecast models were in agreement that the tropical storm would made its way during the following days over the Bay of Bengal into a west-northwesterly direction and heading for the east coast of India. The numerical weather models predicted only a moderate intensification and the system should arrive as a category 1 tropical cyclone named Phailin at the Indian mainland. But on October 10, 2013, the tropical storm strengthened unexpectedly and almost unprecedented rapid into a fully developed category 5 tropical cyclone east of the Andamans.

Figure 1: Images to Track of Super Cyclone Phailin (October 09-13,2013) and the associated storm force(green) and hurricane force (green)winds.

The mean wind speed on October 10, 2013, 00 UTC, was 55 kt (102 kph). Only 24 hours later they had increased to 135 kt (250 kph). Thus, within only one day the tropical storm grew into a category 4 tropical cyclone, which is the second highest category according to the Saffir-Simpson hurricane scale.

With sea surface temperatures between 27 and 30 °C the upper water layers of the Bay of Bengal provided enough latent heat. These values were close to the long term average in that area.

In addition, the wind shear between the upper and lower troposphere was weak enough (about 25 kph) and the tropical cyclone could evolve and keep its vital symmetric structure.

Record low central pressure of Phailin

Phailin showed its maximum intensity between October 11, 12 UTC, and October 12, 00 UTC in the middle of the Bay of Bengal. With maximum 1 min-sustained winds of 140 kt (259 kph) and gusts as strong as 170 kt (315 kph) Phailin was classified as category 5 super cyclone. Phailin equaled the typhoon Usagi, which was previously the world's strongest tropical cyclone of the 2013 season over the western Pacific. According to satellite observations (NOAA) Phailin had a minimum central pressure of 910 hPa on October 10 and 11, one of the lowest pressures ever observed in the territory of the North Indian Ocean. However, these observations are uncertain as buoy measurements and reconnaissance flights are not available in the region for verification. The Joint Typhoon Warning Center ( JTWC) issued a minimum central pressure of 914 hPa, and the number of 918 hPa was given by the Naval Research Laboratory (NRL) .

Landfall of Phailin

Shortly before landfall the tropical cyclone interacted with the India mainland and began to weaken due to the dwindling energy source (warm ocean waters) and increasing friction effects.

On October 12, 2013, the center of Phailin crossed the coastline at 15:45 UTC south of the city of Brahmapur in the Indian federal state of Odisha. At that time Phailin still was a category 4 cyclone. The mean wind speeds taken from satellite observations were about 120 kt (222 kph). The weather station in Gopalpur observed a gust of 185 kph at the storms northern eyewall. Before the weather station failured at 17:10 UTC, a minimum air pressure of 937.4 hPa was measured. After landfall, the former super cyclone weakened rapidly into a category 2 tropical cyclone. Phailin moved into a northerly direction towards the Himalaya and on October 13, the cyclone was identified only as a tropical depression over the North East of India.

Phailin on satellite images

During its maximum intensity Phailin had an enormous extent. On satellite images (Figure 2) the outer cloud bands are spiraling as far as over Sri Lanka and the southern tip of India in the south and over northern Bangla Desh and even the Himalaya at the northern edge of the storm. The storms circulation covered nearly the entire Bay of Bengal and affected an area with roughly 2500 km in diameter. The storm center, the eye, is clearly visible until landfall indicating the symmetrical structure and strength.

Heavy Precipitation, wind, storm surge and wave height

In most cases tropical cyclones are accompanied by heavy precipitation and rain amounts easily in excess of 500 mm. The rain amount and the rain pattern depend on the propagation speed of the storm system, its intensity and extension, and the topography of the affected area. In mountainous and rugged terrain rain might be enhanced to amounts of even more than 1000 mm (e.g. Taiwan, Philippines, Reunion).

Phailin delivered a lot of rain but no exceptional high amounts. In Banki (Odisha) a rain amount of 381 mm fell within 24 hour on October 13. On the same day there were 305 mm at Balimundali (Odisha) and 198 mm in Itchapuram (Andhra Pradesh) 198 mm. On October 9, heavy thunderstorms that were associated with the tropical storm Phailin while crossing the Andamans, brought 336 mm at Maya Bandar within 24 hours.

Figure 3: Accumulated rain amount (October 11-17, 2013, 12 UTC).

(Figure 3) above shows the accumulated rain amount between October 11 and 17 over the eastern Indian Ocean and the western Pacific. During this week, three storm systems left their paths in the rain information was derived from satellite data as gathered by the Tropical Rainfall Measuring Mission (TRMM) of the NASA. Most of Phailins rain fell over the Bay of Bengal. However, along its inland track from the coast of Odisha towards the Himalayan mountains in the Indian federal states of Jharkand and Bihar Phailin released notable rain amounts around 100 to 250 mm. Much more rain was associated with the passage of Nari (landfall in Vietnam) and Wipha. The latter was responsible for Japans sixth- highest rain amount ever recorded within 24 hours: 822 mm in Oshima.

The maximum wind gusts exceeded 300 kph (315 kph) while Phailin was classified as a category 5 super cyclone. At this time the JTWC specified the maximum significant wave height in the open waters of the Bay of Bengal with 54 ft (16 m). Approaching the coastline the cyclone kept its wind gusts well above 200 kph.

Along the coast of Odisha, storm winds piled up a storm surge which penetrated some several hundred meters into the coastal hinterland. According to the Times of India and the BBC, the storm surge was up to 3 meters high, forecasts saw the highest storm surge of about one meter around the town of Gopalpur.

Disaster Profile

History of tropical cyclones over the North Indian Ocean and classification of "Phailin"

Whereas over the west Pacific Ocean usually several category 5 cyclones develop every year, such strong tropical cyclones are much less common over the North Indian Ocean. Phailin was the first super cyclone in the Indian Ocean since 2007 and a maximum mean wind speeds of 140 kt (259 kph) made the storm to one of the strongest ever observed in this area. Only Gonu in 2007 was a stronger cyclone (145 kt, 269 kph).

The last similarly strong tropical cyclone in the Bay of Bengal occurred in late October 1999, when the large 1999 Odisha Cyclone also came along with mean wind speeds of 140 kt (259 kph) and a minimum central pressure of 912 hPa. The 1999 Odisha Cyclone was the first cyclone that was titled as "Super cyclone" by the Indian Meteorological Service (IMD). When making landfall, this cyclone had mean wind speeds of 135 kt (250 kph) exceeding those of Phailin by 15 kt. The Odisha Cyclone fell ashore 160 km further north than Phailin and was accompanied by a 5.9- meter storm surge and caused 9,658 deaths making this storm ranking 4th on the list of deadliest cyclones in India in the last 100 years. Other category 5 tropical cyclones in the North Indian Ocean were Sidr in 2007 and the 1991 Great Bangladesh cyclone. 26 out of the worlds 35 deadliest tropical cyclones raged in the regions around the Bay of Bengal. 42% of all fatalities caused by tropical cyclones refer to Bangla Desh, 27% to India. In November 1977 14,204 people lost their lives, as the Andhra Pradesh Cyclone made landfall just a little bit south of where Phailin hit the Indian mainland. Most devastating was the great Bohla Cyclone in November 1970 that went ashore in Bangladesh (former East Pakistan). The Bohla Cyclone caused a storm surge with a height of more than 10 feet in the Ganges delta and claimed 300,000-

500,000 human lives.

Use of twitter messages for rapid assessment

Figure 4: Number of tweets per hours containing one of the keywords “phailin” and “cyclone” from October 9, 18 UTC, until October 14, 16 UTC.

To get local, detailed, and up-to-date information about the behavior of the cyclone and its impact, Twitter messages (tweets) with various keywords such as cyclone, phailin, shelter, and storm or power outage have been recorded.

(Figure 4) illustrates the number of tweets per hour containing one or both of the keywords “phailin” and “cyclone”. Phailin got much attention from the day before making landfall on October 11, nearly 1,500 tweets were written during the hour of landfall at October 12, 16 UTC, and also during the day after landfall Phailin was very present.

Information Gap Analysis

The chart below (Figure 6) is the result of an analysis of the information produced within the first 4 days following landfall. The information was obtained from ReliefWeb (http://reliefweb.int/disaster/tc-2013-000133-ind), and was retrieved as it was released. All information obtained was categorized under the headings listed on the left side of the graph. Three types of information have been identified (right of the graph) as Basic Data, Analysis, and Root Causes. 'Basic Data' is purely factual and makes up the majority of the information. 'Analysis' consists of information which results from review of this basic data.

It consists of predictions and warnings, as well as identification of levels of needs met or outstanding. 'Root Causes' refers to information which identifies why aspects of the disaster occurred. For example, the low casualties observed in the aftermath of Cyclone Phailin was identified as being the result of a good warning system and excellent coordination between agencies which successfully evacuated almost one million people prior to landfall.

This type of information is very important to disaster risk reduction activities, which attempt to learn from past failures and success by understanding the root causes of each.

The Basic Data is quantified by reviewing how much is produced and how fast each piece of information is provided. Therefore, each of the who, what, where, and when type information, unique to each category, is measured based on how fast it is produced. The Analysis information is also quantified in this way but with a more forgiving time- scale as it will understandably take a little longer to produce. The Root Causes are quantified using only the amount of information, as the timing of this information is not relevant to the immediate disaster response. The dotted line represents the highest potential value of the information produced under each category, being very fast and containing all required information Therefore, the solid bars are percentages of the total potential and the dotted areas indicate where the information gaps are.

4. HARDWARE AND SOFTWARE:

In order to prepare the office for use of GIS mapping capabilities, a new server was purchased with a half a terabyte of storage. Also, 1 gigabit (Gb) network cards were installed on all operational PC’s to allow a faster loading of the large map files onto the PC’s.

The software used for completing the GISbased database included ESRI.s ArcGIS 9 and Microsoft Office Access. It is recommended that SVRGIS data files be viewed using the latest ESRI ArcGIS version. However, free downloadable GIS software applications such as ESRI.s ArcExplorer can be run on the following supporting platforms: Windows, Macintosh, Solaris, AIX, HP-UX, and Linux.

Figure 5: Information Gap Analysis of Super Cyclone Phailin, India Image Credit: CEDIM

5. GIS METHODOLOGY:

The Access database is then imported into ArcGis and becomes a “layer” on the maps that are created for each county. Other GIS “layers” include roads, rivers/streams, towns, reservations, lakes, topography, and rural addressing. A critical layer that has been recently added is “land use.” This layer proved particularly useful in locating residents to provide “ground truth” information on severe weather events. When the program is started, the user opens the file for the desired county. When it loads, the entire county can be seen, along with layers that can be turned on or off including topography (useful for hydrological reasons), towns, roads, rivers and streams, lakes and reservoirs, and property ownership maps. A user begins by choosing an icon from a menu that is shown in Fig. 2. Agnification/demagnification is accomplished through placing the mouse on the image and using the left mouse on the ArcGIS maps with red square areas, as further described below. Before the storm moved into Valley County on 2 July 2005, the operations staff used the ArcGIS maps to find a landowner that the storm had passed; as shown by the land area circled in Fig. 3. The “i” button allowed the forecast staff to retrieve the landowner’s name and several other pieces of information about the land owner. Once this information was obtained, the forecaster was able to retrieve the resident’s phone number from a librar y of phone books, and upon calling, find out the “ground truth” activity in the area. Throughout the severe weather season in 2005, the ArcGIS maps proved invaluable, as WFO experienced the busiest severe weather season in over 50 years of record keeping. This suggests that in many other rural areas across the US, the ArcGIS maps could prove to be a highly useful source for severe weather verification.

6. SUMMARY AND DISCUSSION:

Building on the capabilities of Severe Plot and other graphical interface programs, a GIS severe report graphical and statistical database was constructed to complement existing severe weather graphical databases. SVRGIS is not meant to replace Severe plot and other existing programs, but rather complement these user-friendly programs in situations where more in-depth analysis is necessary. SVRGIS is much more than a tool to query only severe weather because its utility is far-reaching and cross disciplinary. SVRGIS is unique from other graphical databases because it allows the GIS user to import other data types (e.g., - geographical data) and analyze the data by its attributes by means of both spatial and statistical- methods in layer formats. There are countless ways to investigate severe reports and more analysis possibilities exist within a GIS. It is not new to- qualitatively and -quantitatively assess regions of the contiguous U. S. with key elements, such as dollar loss. Population areas at risk et cetera. By using a GIS, it is now possible to assess these variables and many more thorough examination of a plethora of possible relationships between variables all in one database. The possible users of a GIS severe report database include governmental organizations, researchers, academic professionals and students, the emergency management community, and private sector concerns such as insurance companies and private weather firms.

7. REFERENCES:

1. Brooks, H. E., C. A. Doswell III, and M. P. Kay, 2003: Climatological estimates of local daily tornado probability for the United States. Wea. Forecasting, 18, 626-640.

2. Hart, J. A., 1993: SVRPLOT: A new method of assessing and manipulating the NSSFC Severe Weather data base. Preprints, 15th Conf. on Severe Local Storms, Baltimore, MD, Amer. Meteor. Soc., 40- 41.

3. Baumann, P., A Dehmel, P. Furtado, R. Ritsch and N. Widmann (1999), Spatio-Temporal Retrieval with RasDaMan, Morgan Kaufmann, Edinburgh, Scotland, UK, pp. 746–749.

4. Libkin, L., R. Machlin and L. Wong (1996), A query language for multidimensional arrays: Design, implementation, and optimization techniques, Montreal, Canada, pp. 228–239.

5. Marathe, A. and K Salem (1999), Query Processing Techniques for Arrays, Philadelphia, USA, pp. 323–334.

6. Rozumalski, R. (2013), NEWR EMS website. URL: http: // strc. comet. ucar. edu/ software/ newrems

7. Sarawagi, S. and M Stonebraker (1994), Efficient Organization of Large Multidimensional Arrays, Houston, USA, pp. 328–336.

8. ESRI, cited 2005: ArcGIS Site. [Available online at http://www.esri.com/index.html.]

9. Arkansas State GIS Office, cited 2005: The Benefits of Using GIS. [Available online at http://www.gis.state.ar.us/ AGIO_index.htm.]

10. NOAA, cited 2005: NOAA Climate Transition Program (NCTP). [Available online at http://www.ogp.noaa.gov/mpe/n ctp/index.html.]

11. Montana State Library, cited 2005: Montana GIS Data. [Available online at http://nris.state.mt.us/gis/default.htm.]

12. Hart, J. A., 2005: Local severe weather climatologies for WSR-88D radar areas across the United States. Conducted at 2005 National Weather the weather because its utility is far-reaching and cross disciplinary. SVRGIS is unique from other graphical Associations 30 Annual Meeting, St. Louis, MO. databases because it allows the GIS user to import other datatypes (e.g., - geographical data) and analyze the data by its attributes by means of both spatial and statistical- methods

13. Hart, J. A., 2006: .Online severe weather database: gridded analyses. Online posting. 6 Mar. 2006. NOAA/NCEP/Storm Prediction Center. 29 Apr. 2006.<http://www.spc.noaa.gov/climo/online/grids/index.html>

14. Schaefer, J. T. and R. Edwards, 1999: The SPC tornado/severe thunderstorm database. Preprints, 11th Conf. On Applied Climatology, Dallas, TX, Amer. Meteor. Soc., 215-22

Received on 20.10.2015 Accepted on 30.10.2015

Research J. Engineering and Tech. 6(4): Oct. - Dec., 2015 page 432-438

DOI: 10.5958/2321-581X.2015.00067.7