Exploratory Data Analysis of Damages in Hurricane Harvey (2017 USA) with MATLAB
Exploratory Data Analysis with MATLAB
Hurricane Harvey
Hurricane Harvey was a devastating Category 4 hurricane that made landfall on Texas and Louisiana in August 2017, causing catastrophic flooding and more than 100 deaths. It is tied with 2005's Hurricane Katrina as the costliest tropical cyclone on record, inflicting $125 billion (2017 USD) in damage, primarily from catastrophic rainfall-triggered flooding in the Houston metropolitan area and Southeast Texas; This made the storm the costliest natural disaster recorded in Texas at the time. It was the first major hurricane to make landfall in the United States since Wilma in 2005, ending a record 12-year span in which no hurricanes made landfall at the intensity of a major hurricane throughout the country. In four days, many areas received more than 40 inches (1,000 mm) of rain as the system slowly meandered over eastern Texas and adjacent waters, causing unprecedented flooding. With peak accumulations of 60.58 in (1,539 mm), in Nederland, Texas, Harvey was the wettest tropical cyclone on record in the United States. The resulting floods inundated hundreds of thousands of homes, which displaced more than 30,000 people and prompted more than 17,000 rescues.
read more about Hurricane Harvey on Wikipedia
Background
MATLAB Programming
Importing the data
StormEvents_2017 = importStormFile("StormEvents_2017_finalProject.csv");
Harvey = StormEvents_2017(ismember(StormEvents_2017.State,{'ARKANSAS','KENTUCKY', ...
'LOUISIANA','MISSISSIPPI','NORTH CAROLINA','TENNESSEE','TEXAS','ALABAMA', ...
'FLORIDA','GEORGIA','SOUTH CAROLINA'}),:);
Since Hurricane Harvey-related events occurred only from the 17th of August to the 3rd of September, we are analyzing the data between these two dates, as shown in the Table.
Harvey = Harvey(ismember(Harvey.Month,{'August','September'}),:);
Harvey = sortrows(Harvey,'Begin_Date_Time');
Harvey = Harvey(Harvey.Begin_Date_Time >= '17-Aug-17 00:00:00' & ...
Harvey.Begin_Date_Time <= '03-Sep-17 00:00:01',:);
Harvey(1:5:30,3:1:9)
The storm events from 17th August to the 3rd of September can be seen in the following density plot.
geodensityplot(Harvey.Begin_Lat,Harvey.Begin_Lon)
legend('storm events')
title("Hurricane Harvey events from 17th August to the 3rd of September")
Two states most impacted by Hurricane Harvey
The two states most impacted by Hurricane Harvey in terms of total Property Cost estimated. 275 and 86 storm events occurred in the states of Texas and Louisiana, respectively. The Sum of property costs in these two states is $77427 Million and $75.277 Million for the states of Texas and Louisiana, respectively.
total_cost=groupsummary(Harvey,"State","sum","Property_Cost");
total_cost = sortrows(total_cost,'sum_Property_Cost','descend');
Impacted_states=total_cost(1:2,:)
Table of Events for Two Most Impacted States
A few rows of the table of the two most impacted states are shown below. This table includes only the states of Texas and Louisiana.
two_states = Harvey(ismember(Harvey.State,{'LOUISIANA','TEXAS'}),:);
two_states = sortrows(two_states,'Deaths_Direct','descend');
head(two_states)
Visualizations
The figure of Event Types
The following Histogram shows the different types of events that occurred during Hurricane Harvey in the most impacted states of Texas and Louisiana. The most frequent event was Flash Floods which occurred 179 times.
%filtering the event types which didn't occure, or their frequency is 0.
two_states.Event_Type = removecats(two_states.Event_Type);
histogram(two_states.Event_Type)
grid on
title('Histogram showing the frequency of different events')
xlabel('Event Type')
ylabel('Number of occurance')
The following heatmap shows a count of Event types vs the States of Texas and Louisiana.
two_states.State=removecats(two_states.State);
heatmap(two_states,"State","Event_Type")
The figure of Event locations
following is the geoscatter plot to show the storm event locations in the states of Texas and Louisiana.
texas=two_states.State=="TEXAS";
geoscatter(two_states.Begin_Lat(texas),two_states.Begin_Lon(texas))
geolimits([25.6 37.4],[-106.09 -88.60])
hold on
lsiana=two_states.State=="LOUISIANA";
geoscatter(two_states.Begin_Lat(lsiana),two_states.Begin_Lon(lsiana))
legend({'Events in Texas State','Events in Louisiana State'})
hold off
title('Figure showing the locations of events in states of Texas and Louisiana')
Analysis
Three Counties with Most Events in Texas State
The three counties with the most events in Texas are shown in the following table. The county Harris had the largest number of events which is 21.
state_1 = two_states(two_states.State == 'TEXAS',:);
state1_3count=groupcounts(state_1,"CZ_Name");
state1_3count = sortrows(state1_3count,'GroupCount','descend');
state1_3count=state1_3count(1:3,:)
Three Counties with Most Events in Louisiana State
Natchitoches, Sabine, and Red River are the counties with the most events in the state of Louisiana
state_2 = two_states(two_states.State == 'LOUISIANA',:);
state2_3count=groupcounts(state_2,"CZ_Name");
state2_3count = sortrows(state2_3count,'GroupCount','descend');
state2_3count=state2_3count(1:3,:)
Three Counties with Highest Property Cost in Texas State
The following table shows 3 counties with the highest property costs in Texas state. These counties are Galveston, Fort Bend, and Montgomery with property costs of $20, $16.004, and $14 Billion, respectively. It is assumed that these costs are estimates from models.
Hproperty_s1=groupsummary(state_1,"CZ_Name","sum","Property_Cost");
Hproperty_s1 = sortrows(Hproperty_s1,'sum_Property_Cost','descend');
Hprop_s1c3=Hproperty_s1(1:3,:)
Three Counties with Highest Property Cost in Louisiana State
The following table shows 3 counties with the highest property costs in the Louisiana state. These counties are Calcasieu, Beauregard, and Acadia with property costs of $60, $15, and $0.2 Million, respectively. It is assumed that these costs are estimates from models.
Hproperty_s2=groupsummary(state_2,"CZ_Name","sum","Property_Cost");
Hproperty_s2 = sortrows(Hproperty_s2,'sum_Property_Cost','descend');
Hprop_s2c3=Hproperty_s2(1:3,:)
Conclusions and Recommendations
Texas State
After analyzing the storm data, we know that the state of Texas is the most affected. a total of 275 events occurred in the state during hurricane Harvey. The number and types of events can be seen in the histogram in the visualization section. Most of the events occurred in the Harris, Galveston, and Fort Bend counties. In terms of property cost, Texas state has a property cost of around $77 billion. Galveston, Fort Bend, and Montgomery are the three counties in the Texas state with the highest property cost of $20, $16.004, and $14 Billion, respectively
Louisiana State
The second most impacted state is the state of Louisiana. around 86 events occurred in this state. Most of the events occurred in the Natchitoches, Sabine, and Red River counties. In terms of property cost, this state has a property cost of around $75 million. Calcasieu, Beauregard, and Acadia are the three counties in the Louisiana state with the highest property cost of $60, $15, and $0.2 Million, respectively.
Recommendations
It is recommended that the priority for the insurance company should be the states of Texas and Louisiana, with the state of Texas at the top. The company should send most outside contractors to the Texas state with priority in the counties of Galveston, Fort Bend, and Montgomery. For the state of Louisiana, the company should send outside contractors to the counties of Calcasieu, Beauregard, and Acadia.
Comments
Post a Comment