Investigating Historical Tornadoes Using ArcGIS Name: Summary: In this lesson, you will use ArcGIS (version 10) and spatial analysis to investigate historical tornadoes across the USA. Objectives: Work with data and maps to understand the spatial and temporal distribution of tornadoes in the USA across multiple decades. Software: This lesson uses ArcGIS Desktop 10 software, from Esri: http://www.esri.com/arcgis. Earlier versions of the software could be used but the best results will be achieved with ArcGIS 10. Analyzing historical tornadoes using GIS and the spatial perspective in this lesson. Level: This activity can be used at a variety of levels from upper secondary to university level. Questions could be modified depending on the level targeted. Students should be familiar with maps and some basics of physical and cultural geography and some GIS to most effectively use this lesson. Time Required: This lesson includes 70 questions. Estimated time to complete them is 4 standard class periods. Or, it could be assigned outside of class with an estimated time of completion as 2 to 3 hours. The investigation could be extended to one week of 5 class periods, allowing for class presentations to be made. Activity: These questions and problems can be posed by an instructor in a onecomputer classroom with a projector, engaging the students in discussion while using the ArcGIS tools, or in a computer lab setting where students are working on their own computers. In a lab setting, students can work independently or in groups through these problems. Skills: This lesson includes analyzing spatial point and polygon data, symbolizing data, data and file management, selection, sorting, and querying data in maps and tables, working with map projections, creating summary tables, and creating and analyzing data in graphs. Extension: These lab sheets are all you need to run the lesson. However, you could easily extend the lesson by investigating more recent tornadoes from NOAA, or using an elevation data set for a different region of the USA, or examination of cities and their proximity to tornadoes, or many other data sets, depending on the
questions you wish to ask. You could also generate a tornado density and a tornado 3D surface using Spatial Analyst and 3D Analyst, respectively. If you extend the lesson, please consider submitting your extension as a lesson on ArcLessons (http://edcommunity.esri.com). Activity In this activity, you will examine historical tornado touchdown points and tracks in the USA using ArcGIS Desktop 10 software. Data Management Copy the data folder to your computer or on the network where you will store the data. A copy of this exercise and the data you need are in a folder called tornadoes in the P:/gis/Physical_Geography folder. A copy of this exercise is MS Word format as well as.pdf Data is found in the P:/gis/Physical_Geography/tornadoes folder Open the file tornado.mxd in ArcMap. You now should see the following 4 data layers in your map: Elevcent elevation for the central region of the USA States state outlines Tornadoes 1950-2004 tornadoes for the USA Tornado Tracks 1950 2008 for the USA Problem: You are a scientist with the National Oceanic and Atmospheric Administration s (NOAA) National Climatic Data Center; http://www.ncdc.noaa.gov. NOAA is asking you to assess tornado touchdowns and tracks so that you can better plan your staff and resources to handle these types of events. Included in your assessment will be an analysis of patterns, by month, by severity, by state, by elevation, by specific city, and tornado touchdowns versus tracks. Analyzing Tornado Patterns 1) How many tornado touchdowns are in the data set? 2) What is the time period that the tornado touchdown data set covers? 3) State 3 observations about the pattern of tornados that across the USA. 4) Name 2 things about the pattern of tornados that across the USA that surprises you.
The current symbology indicates the intensity of the tornados. There are so many tornadoes that it is difficult to determine where the most severe (F_SCALE 4 and 5) tornadoes occur. Therefore, perform a Definition Query on the tornado touchdowns (right click on layer Properties Definition Query) to reduce the data shown to F_SCALE of 4 and 5, as follows: 5) What do you observe about the pattern of the intensity of the most severe tornados? 6) Is the pattern of the most severe tornadoes different from the pattern of all tornadoes? If so, how? 7) Access the table and compare the selected number of tornadoes to the total number of tornadoes in the data set. What percent of all tornadoes in the data set have a Fujita of 4 and 5? Show your work. 8) Go to properties symbology for the tornado touchdowns layer. If no numbers appear under the count field, click on it to calculate these numbers. What percent of all tornadoes have no data (-9 or 0) for the Fujita scale?
9) Think about the time period that your data covers. Why do you suppose that so many tornadoes are missing the values for the Fujita scale? Change the legend and/or run definition queries to answer the following questions. 10) What do you observe about the pattern of the tornadoes with the longest lengths? 11) What do you observe about the pattern of the tornadoes with the longest widths? Analyzing Tornadoes by Fatalities and Injuries Highlight the fatalities (FATAL) field and sort the data on this field. Select the record in the table where the most fatalities occurred. Exit the table and return to the map. Use the "zoom to selected" function to center this tornado on the map. Use the zoom into center and identify tools to answer the following questions. 12) In what state did the tornado causing the most fatalities occur? 13) How many fatalities occurred? 14) When did the tornado occur? Clear your selection and repeat your process above to answer the following questions about the tornado with the most injuries. 15) What state (field STATE2DIG) did this tornado occur in? 16) How many injuries occurred? How many fatalities occurred? 17) When did the tornado occur? Analyzing Tornadoes by Location You can also create new data in ArcGIS. This data may be a summary set of data, as you will do next. Access the table for the tornado touchdowns and make state the selected field. Summarize the field. Open the new summary table and sort the summary table on count. 18) Which state is listed as having the most tornadoes? Label the states on your map.
19) Which state had the fewest tornados in the list but still had some tornados? 20) Did any state have no tornadoes? If so, which one(s) were they? 21) To which 5 states will you recommend that NOAA direct the most tornado relief and research funds? 22) To which 5 states will you recommend that NOAA direct the least tornado relief and research funds? Analyzing Tornadoes By Time Use the pan and zoom functions in the view, table sort function, identify button, select by attributes, and the definition query tools that you used previously in this lesson to answer the following questions. 23) During which month did more tornados occur than any other month? 24) During which month did fewer tornados occur than any other month? 25) During which month did the extent of tornados across the USA reach what you observe to be the maximum spatial extent? Conduct some research on the causes and pattern of tornadoes in the USA. 26) Why do you think the spatial maximum occurs after the numeric maximum? 27) Conduct some research on tornadoes worldwide. Give two reasons why most of the world s tornadoes occur in the USA. 28) Create a layout showing all April tornadoes throughout the USA. Include your name, a legend, and a scale. Using GIS to Analyze a News Story In August 1999, during the Outdoor Retailers conference where the USGS (where the author worked at the time) and other organizations were operating exhibits, a tornado raced through downtown Salt Lake City, Utah, killing one person and devastating parts of the downtown urban area. The newspaper reported that until that day, no fatalities from a tornado had ever occurred in Utah. 29) According to the data set that you are analyzing, was this newspaper report correct? 30) Have there ever been any injuries due to tornadoes in Utah? If so, what years did injuries occur?
Analyzing Tornadoes by Elevation Next, you will examine the relationship of elevation and tornado occurrence for the Central USA. Make the layer elevcent visible. 31) Describe the elevation in the Central USA. 32) Are the elevations in the legend in feet or meters? How do you know? 33) What is the name of the field in elevcent that contains the elevation category? 34) Circle the correct answer below: [ A ] A higher elevation category corresponds to a higher elevation. [ B ] A higher elevation category corresponds to a lower elevation. 35) Make a visual assessment of the number of tornadoes across the elevation zones. Your assessment for NOAA needs to include more than a visual assessment, but quantitative data. Fortunately, you have the data and GIS tools at your fingertips, and will be able to provide it after performing a few functions. 36) Examine the field names for the tornado touchdowns. Do you have information that tells you how many tornados occurred in each elevation category? Therefore, you need to join the elevation information from the elevation layer to the tornado data so that you can determine which tornados were inside each elevation class. You will do this in 3 steps. First, reduce the tornado data set that you will analyze to just those tornadoes in the central region, by using the Select By Location function on the tornado touchdowns and elevcent layers, as follows:
37) How many tornadoes occurred within the central region? Second, save your selected tornadoes for the central region as a separate feature class in your tornadoes geodatabase by right clicking on the tornado touchdowns layer and saving it as torncent. Third, right click on the central region tornadoes layer torncent and access Joins and Relates. On the drop-down, select join data from another layer based on spatial location, as follows. Join tornados to the elevcent layer. Take the defaults in #2 and in #3, give the output a meaningful name, such as tornelevcent, saving it as a feature class in your geodatabase. Be patient. The GIS has to assign an elevation to all tornadoes in the central USA. Clear the selected set of tornadoes from the tornado touchdowns layer. Make this layer invisible.
38) When complete, examine the field names in the attribute table for tornelevscent. What new field(s) have been added? In this attribute table for tornelevcent, select the field gridcode and summarize it. Save the summary table to your folder. Give it an appropriate name. Add the result to your map document. Sort the summary table on count_gridcode. 39) Which elevation category had the most tornados? What is the elevation of this category? 40) Is this what you expected? Why or why not? In the summary table, create a graph of the distribution of tornados by the count of the gridcode. Create a bar graph that includes your name and an appropriate title, something that looks similar to the graph below: Include your graph as a screenshot below. 41) What does this graph tell you? Analyzing Risk to Specific Cities from Tornadoes In this section, you will once again turn your attention to analyzing all tornadoes, not just those in the central region. Therefore, turn on the tornado touchdowns layer and turn off the tornadoes and elevations associated with the central region.
Under the Add Data button, select Add Basemap. This will access a set of basemap choices from ArcGIS Online. Select DeLorme. Wait a moment while the basemap draws in your projected geographic space (you may receive a projection warning). Now you have access to a basemap. But because that basemap is online, you may experience some delays in the draw speed. When using the basemap, you may wish to toggle it off if you are zooming and panning a great deal, and then turn it back on. Access View Data Frame Properties, and set the display units to miles, as follows. You may receive a warning.
Access the select by circle tool, as shown below: Pan the map to Oklahoma City. Click on the center of Oklahoma City. Keeping an eye on the radius of the circle as you draw, drag the circle outward until the radius is as close to 50 miles as you can make it. Access the attribute table of tornado touchdowns. 42) How many tornadoes occurred within 50 miles of Oklahoma City? Run statistics on the field INJ on the selected tornadoes. Look at the sum of INJ (injuries). 43) How many injuries occurred from these tornadoes? 44) Perform the same assessments on the following cities and record the data. City Denver, Colorado Number of tornadoes within 50 miles Number of Injuries Kansas City, Missouri Compare the results above to that for Oklahoma City. 45) Which of the 3 cities has experienced the most tornadoes within 50 miles of its downtown? Which of the 3 cities has experienced the least tornadoes within 50 miles of its downtown? 46) Which of the 3 cities has experienced the most injuries from tornadoes within 50 miles of its downtown? Which of the 3 cities has experienced the least injuries from tornadoes within 50 miles of its downtown? 47) Can you find another major city where more tornadoes occurred within 50 miles of its downtown than the any of the cities you tested above? If so, what city is it and where is it?
Analyzing Tornado Tracks Tornadoes do not simply touch down, of course. Part of the reason they are so destructive is that they move across the land. Toggle the tornado touchdown layer and the tornado tracks layer. 48) Compare and contrast the tornado touchdown layer with the tornado tracks layer, naming 2 similarities and 2 differences. 49) Which layer the tornado touchdowns or the tornado tracks do you believe gives you a better sense of the pattern for tornadoes across the USA? Why? 50) What direction would you say is the predominant direction that tornadoes move across the landscape in the USA? From to. Use your research on tornadoes to answer the following question. 51) Why do tornadoes move in this direction across the landscape in the USA? Symbolize the tornado tracks as graduated symbol by length, with the longest tornadoes shown with the brightest color. 52) Consider the pattern of tornado tracks by length in the USA. Is there a pattern? If so, what is it? Use the measure tool and the scale bar to answer the following question. 53) What do you think is the unit of measure for the length of the tornadoes in this data set? Symbolize the tornado tracks as graduated symbol by width, with the widest tornadoes shown with the thickest line.
54) Consider the pattern of tornado tracks by width in the USA. Is there a pattern? In other words, do wider tornadoes occur in certain locations more than others? If there is a pattern, what is it? 55) Use the measure tool and the scale bar to answer the following: What do you think is the unit of measure for the width of the tornadoes in this data set? 56) Why is it important to know the units of measure when doing GIS analysis? Examine the Year field. following question. Use the Summarize function on year to answer the 57) According to this data set, are tornadoes becoming more or less common as time goes on? 58) Think about how tornado data was collected during the 1950s and 1960s versus today. Before radar and other remote sensing techniques existed, how was tornado location data gathered? 59) How does the method of data collection impact the frequency of data in any data set, tornado or otherwise? How do these considerations impact your assessment whether tornadoes are becoming more frequent? Examine the fatalities, injuries, Fujita, and damage fields. 60) Can you detect any relationship between the Fujita scale and fatalities/injuries? If so, what is it? 61) Can you detect any relationship between the Fujita scale and damage? If so, what is it? The HourLocal field shows, on the 24-hour clock, the time that the Use the summarize function on the field HourLocal to answer the following question. 62) List the three most frequent hours of the day when tornadoes were most common, according to this dataset, in descending order beginning with the most frequent hour in which tornadoes occurred. 63) Thinking about how and why tornadoes form, why do you think these three hours are the most common for the occurrence of tornadoes? Examine the date field. 64) Can you find a date where more than 10 tornadoes occurred? If so, when was it? 65) Why do tornadoes sometimes occur in groups?
Select one of the dates where more than 10 tornadoes occurred. 66) Describe the geographic distribution of these tornadoes. Find the Select By Polygon tool as shown at right: 67) Measure the geographic area (in square miles) that these tornadoes occurred and indicate it below. square miles. Synthesis 68) Summarize in a paragraph or two what you have learned about tornadoes from this exercise 69) Summarize in a few sentences what you have learned about GIS and spatial analysis in this lesson. 70) Summarize in a few sentences what you have learned about the pattern of tornadoes in this lesson.