In our study, this new method was applied, and the geographic correlation of the lightning damage with several environmental and socio-economic variables were examined. Lightning strikes that caused damage from January 1993 to December 1995 in Mecklenburg County, North Carolina were geocoded and the possible areas where lightning strikes could occur were located, and the geographical pattern of lightning damage was studied. In addition, the statistical correlation between the cost of the lightning damage and different variables was analized.
In this exploratory study, we reviewed the cost of lightning strikes that occurred in Mecklenburg County, North Carolina from January 1993 to December 1995. We were trying to find out the geographic distribution of the cost of lightning damage and whether any environmental or socio-economic variable had significant effect on the cost of lightning damage. If we could find the pattern of geographic distribution of the cost of lightning damage or significant correlation between certain variables with the cost of lightning damage, we could predict the potential high damage area of lightning strikes. It would be very helpful in future planning of this area, and it will be helpful in avoiding the high damages from lighting strikes.
GIS was applied in this study. GIS is a newly developed computer software which can capture, store, manage, extract and display geographically referenced information. It provide statistical summaries, calculations, interrelationships of data, buffer generation and overlay functions. In our study, we used ArcInfo -a major software in GIS. The geocoding of the Lightning strike, locating of the exact area of lightning strikes, integrating of the environmental variables and socio-economic variables with the cost of lightning damage, and finding the geographic pattern of the cost of lightning damage were all accomplished by the functions in ArcInfo.
The geographic pattern of the lightning damage was studied and we found most of the high cost of lightning damage occurred in the South Planning District in Mecklenburg County. Correlation and regression analysis were performed in testing the interrelationships of the environmental and socio-economic variables with the cost of lighting damage. The environmental variables included soil types, pH value, water capacity and slope. The socio-economic variables included the age of the building and the property value.
This study demonstrated the ability of GIS method to locate the possible area of the lighting strikes. This enables the linking of the geographic and socio-economic variables with the cost of lightning damage and makes the study of the causes of lightning damage feasible.
The environmental variables included soil type, pH value, water capacity and slope. They came from the Soil Conservation Survey (SCS). The socio-economic variables included the age of the building and the property value. They were obtained from Zoning and Addressing Department in Mecklenburg County.
In this study, several other GIS data layers were used. Streets centerline file in 1:100,000 scale updated in June 1996 was used in the geocoding and locating of the area of lightning damage. The Planimetric layer in scale 1:2,400 obtained from a 1992 airphoto was used in the locating of the lightning stricken area. Soil coverage in scale 1:24,000 was used in the integration of data. They all were obtained from the Engineering Department of Mecklenburg County.
1. The first step of the data base development was to find the possible area for each lightning strike that caused damage. There are two types of lightning damage: One is caused by direct hits from lightning strikes, which means lightning directly hits a building and causes damage; the other is caused by indirect hits, such as lightning hits a tree and travels into a house through a power line and causes damage. The lighting damage data obtained in this study were all 911 calls and had the exact address of each building. Since no records were obtained to show whether these damage were from direct hits or indirect hits of lighting strikes, all the possible area that lightning strike could occur should be included in our study. And it could be an polygon area surrounding each building. In this study, the following steps was performed to define the area of lightning strikes.
First, the lightning strikes were geocoded through the ADDRESSMATCH function. The streets centerline file was used and a point coverage was generated. This point layer contains the information of the address and cost of lightning damage. About 99% percent of the ninety-five records were matched.
Secondly, the exact location of the building was found for each lightning strike. This was accomplished by using the Planimetric layer and the tax parcel information. Since the data we studied were for 1993 to 1995, we found several buildings missed from the Planimetric layer (produced in 1992 ) . Using the streets centerline file as a base map, we manually added the polygons to the sites where the tax parcel information indicated. After this step, a polygon coverage containing all the building roof areas of lightning strikes was generated.
Thirdly, a 1,000 feet buffer was generated around each building. This created a study area which covered all the possible sites that lightning strikes could occur. This polygon coverage was used as study area in the our analysis.
2. In the second step of database development, all the environmental and socio-economic variables that we wanted to test was integrated with the cost of lightning damage. First, the address and cost of lighting damage was added to the PAT file of the buffered study area from the point coverage of lightning damage from the former step by OVERLAY and JOINITEM function. Secondly, the overlay of the study area with the soil coverage was performed and the information of the type of soil was joined into the PAT file of the output coverage. By ADDITEM and JOIITEM function, soil types, pH value, water capacity, the age of each building and property value of the parcel were also added into the same PAT file. This PAT file also contained the addresses and cost of lightning damages. At this point, we integrated all the variables that we wanted to test with the cost of lightning damage, and we were ready to do the analysis.
1. Slight Damage: $10 to $1,000,
2. Minor Damage: $1,000 to $2,500,
3. Moderate Damage: $2,500 to $10,000,
4. Major Damage: $10,000 to $45,000,
5. Severe Damage: $45,000 to $165,000.
The study of environmental variables shows that the 62% of the slope was 5o , 56% of the pH value was 5.1, 67% of the water capacity ranged between 0.14 to 0.15 and about 62% of all the soils was CeB2 or CeD2. The result of the socio-economic variables showed that 90% of all the building was built after 1950 and 50% of the property value was over $100,000, while comparing with the 22% of the cost of lightning damage was over $100,000. This illustrated the severity of lightning damage. Correlation and regression analysis were performed for these factors with the cost of lightning damage in SYSTAT but no significant result was found.
There are some limitations of this study. First, there was no significant correlation found between the cost of lightning damage and the variables we chose in this study. One reason for this could be the limitation of lightning damage data. The data that were available at the beginning of our study were for 1993 to 1995 and contained 95 cases. There is a possibility that these data could not show the statistical correlation. Secondly, some of the data were not available at the time of our study. All these limitations could affect the statistical results.
This study showed the feasibility and capability of applying GIS to the study of Lightning Damage. In testing the geographic distribution of the lightning damage, we found some geographic pattern of lightning damages. However, this study is still at preliminary stage. More variables such as the sea level, building height, aspect of slope and construction of buildings should be analyzed. We prospect this study will lead more thorough and intensive studies of this field.
Jack Horan. " Some residents fear they're in line of fire. " Charlotte Observer, August 17, 1996.
Eric S. Livingston, John W. Nielsen-Gammon, and Richard E. Orville. " A climatology, synoptic assessment, and thermodynamic evaluation for cloud-to-ground lightning in georgia: a study for the 1996 summer olympics " . Bulletin of the American Meteorological Society. vol.77, No. 7, July 1996.
Mecklenburg Planning Commission. " South District Plan ", June 1, 1992. Environmental System Research Institute, Inc. " Understanding GIS", 1995.
Wei-Ning Xiang
Associate Professor, Department of Geography and Earth Sciences
University of North Carolina at Charlotte
9201 University City Blvd.
Charlotte, NC 28223-0001
Telephone:(704) 547-4247
Fax: (704) 547-3182
E-mail:wxiang@uncc.edu
Joseph C. Wilson
Undergraduate Student, Department of Geography and Earth Sciences
University of North Carolina at Charlotte
9201 University City Blvd.
Charlotte, NC 28223-0001
Telephone:(704) 547-4247
Fax: (704) 547-3182
E-mail:jwilson@uncc.edu