Wednesday, March 16, 2011

Final Project- Hurricanes



Introduction
Hurricanes are some of the most intense forces of nature, capable of generating wind speeds of over 150 miles per hour and causing great amounts of destruction.  In 2005, Hurricane Katrina caused levees to break in New Orleans, which resulted in city flooding and hundreds of lost lives. In 1992, Hurricane Andrew swept across southern Florida and demolished entire neighborhoods. Thousands were left homeless. Seeing the impact of such mega-storms, I decided to map out hurricane proneness across the United States in order to identify the areas of highest hurricane risk. Those areas which typically endure hurricanes more than others should be examined with a discerning eye to ensure that evacuation procedures are reliable, adequate supplies will be available, and so on. These maps can be useful in deciding where to move. Moving into an area of higher hurricane risk means buying window shutters and having them be readily accessible. One afternoon in Fall, residents may be asked to board up their homes and evacuate, possibly to never see them again.

Methods
I made maps showing the hurricane-proneness of states and counties. First I made the map to see what the most hurricane-prone states were, and those were the states I chose to analyze hurricane threat on at the county level. I found a shapefile of all previous Atlantic hurricanes from 1851 to 2008. This shapefile was comprised of many individual line segments however, as opposed to pre-assembled full linear courses of the hurricanes. This caused me to do a lot of different types of selections and categorizations to be able to show the hurricane paths based on their category ranking on the Saffir-Simpson Hurricane Scale. This scale goes from 1 to 5, and is used to categorize hurricanes based on their wind speed. Category 1 hurricanes have wind speeds between 74 and 95 mph, while category 5 hurricanes have winds of over 156 mph. Categorizing the storms by this scale was tedious because each small line segment had its own wind speed value of the storm at that particular point along its path. For instance, a category 5 hurricane will have category 1 and category 2 line segment attributes. I found that the most efficient way to accomplish my task was to group together the storms by their identification numbers to ensure they would be displayed as a whole. I plotted every Atlantic hurricane that occurred from 1988 to 2008- twenty years worth- in order to be able to get an idea of where the most common places for hurricanes to strike are over time. In order for a state or county to be considered to have had a direct impact by a hurricane, the hurricane must have either intersected, or have come within 29 miles of the specific polygon. Twenty-nine miles is the average radius of maximum wind, which is located around the eye wall of a hurricane. Although damage can definitely occur, and people hundreds of miles away from the eye of the hurricane can feel and observe its effects, this analysis focuses on those generally impacted the greatest, the places within the radius of maximum wind. I calculated the total number of hurricanes that intersected or came within 29 miles of states over the twenty years. After I got my values per state, I divided the number by 20 in order to provide the average number of hurricanes a place has per year. Most states have a decimal below one, but the two highest, North Carolina and Florida, have averages of 1.3 and 1.0 hurricanes respectively per season. These are the two states I analyzed county hurricane proneness for. County analysis was similar to state analysis, as I would click each county one by one to select it, and perform a select by location based off of it to see how many hurricanes have at one point moved through that spot.

Results
As I mentioned, North Carolina and Florida had the greatest amount of hurricanes come within 29 miles of them as opposed to any other states from the years of 1988 to 2008. All of the states bordering the Atlantic Ocean or the Gulf of Mexico showed hurricane activity. Most of the western states show none. There are generally higher concentrations of more hurricane-frequent states in the south as opposed to the north, which is correlated to a hurricane’s need for warm waters to perpetuate it. Mapping the counties of North Carolina and Florida made me very clear on the effect proximity to the Ocean has of hurricane frequency. North Carolina looks as though it is divided in half by the symbolic colors representing frequency of hurricanes. The western, inland portion of the state has very little hurricane activity compared to that of the coastal region. The average number of hurricanes a western North Carolina county experienced during the twenty years was about two or three, and near the coastal region that number is more along the range of nine or ten. It’s also interesting to look at the types of storms that affected different regions and mind trends in that regard as well. For instance, there is a large distinct band of category 5 hurricanes that have gone over the southern tip of Florida. I also realized the general horse-shoe trend of hurricane movement, in which most form in the middle of the Atlantic at tropical latitudes, move northwest towards the United States, and then turn northeast to go back out into the ocean.

Conclusion/discussion
It is very important to know about places that have higher risks for things than others. Knowing a place’s hurricane proneness is no exception. People in these places must be very aware of their surroundings and may need to act in the case of an emergency. It is interesting to see this distribution of hurricane paths and frequency across states and relating them to other topics. For example, Palm Beach County, Florida, one of the wealthiest places on the planet, is tied for the highest hurricane frequency in the state. It shares its tie with Monroe County, Florida, which bears the Florida Keys, deemed by many as one of the most relaxing places on Earth.  These places are in tropical, coastal, humid environments, the same environments that help bring hurricanes to life. We must be able to prepare for and understand the environments in which we live.

Wednesday, March 2, 2011

Week 8- Interpolation


I chose to interpolate the rainfall data using the Inverse Distance Weighted (IDW) and Kriging methods. It was very interesting mapping the season to date rainfall and the normal amount of rainfall in a given season. This provides for a very telling visual, giving us an idea of where we are in the season, relatively how much rainfall we are recieving, and the places throughout the county that receive more or less rainfall than others. Seeing such data visualized, there are clearly regions of the county that receive more precipitation than others.

This year we have had a very high amount of rain. This is clear on the east side of the difference images, as the greens and yellows represent negative values, meaning that the average seasonal amount of rainfall has already been surpassed during this season. I included a hillshade model in my maps, and it is fairly clear how the rainfall patterns coincide with elevation areas. The eastern part of the county that receives the most rain is in the mountains and the driest part is in the north, above the Grapevine, forming the lower portion of the Central Valley.

Kriging seems like the best way to interpolate this kind of data. It shows very general trends and gradations, which is what is needed for portraying a continuous feature like rainfall across a given area. IDW is good too, but the data is not dense enough for it to be as accurate. It portrays a lot more concentrated areas of rainfall around particular gauging stations, which is not necessarily the realistic case.

Wednesday, February 23, 2011

Week 7- Fire Hazard Mapping

Going through the fire hazard mapping tutorial, things were okay. But when I began mapping the Station Fire hazards on my own, I started to run into some challenges. First off, finding some of the data was not easy. The fire perimeters and vegetation data were particularly hard to find. After I found all the data I needed, I began to do work on the mapping. For some reason, the computer was not allowing me to make a slope map from the DEM data, and an error would come up every time I tried. Soon, my computer froze and ArcMap stopped responding entirely. I opened up another ArcMap window on the same computer and began remaking my map.

After it wasn't allowing me to create a slope map again, I tried using other Los Angeles County DEM data, but ArcMap placed the layer far north of the rest of the Los Angeles layers, roughly in the area of Yosemite National Park. I thought that it may have something to do with the coordinate systems being off, so I tried to open ArcCatalog to fix it. ArcCatalog wouldn't open, as a message came up saying I don't have the licensing to run the software. I switched computers to try and get it to work. This time, neither ArcMap or ArcCatalog would open. I switched computers, then switched labs, and then switched computers again, and the programs would not open anywhere. I talked to someone at the front desk and he told me they were doing maintenance on the software and it should be back up shortly.

After a few minutes, it was up and running again, and this time I was able to create the slope map with the original DEM data. In the vegetation data I acquired, there was a field called fuel ranking, indicating the proneness a particular type of vegetation has to catching on fire. The vegetation hazard map is symbolized accordingly. When I began formatting my map, I added one legend per data frame, but I was frustrated that I was adding a lot of the same information over and over, as the fire perimeters, urban area, and highways are symbolized the same on each map. The only things that are different are the different types of hazards and gradients. I then had the idea to activate all of the layers in one data frame, in order to create one large, encompassing legend.

Wednesday, February 16, 2011

Week 6- Suitability Analysis


There is an issue in Kettleman City, California, a small city in the heart of the state's agricultural Central Valley, in which babies are being born with birth defects. Many believe that this is due to the higher-than-normal amounts of arsenic in the drinking water in the area, caused by a nearby landfill. The fact that the landfill received over 400,000 tons of hazardous waste, including PCBs, in 2009, could easily lead someone to make the assumption that the landfill is a contributor to the birth defects. However, this is also in the middle of one of the largest agricultural regions in the country and pesticides may play a role in the birth defects. Either way, because of these concerns, Senator Diane Feinstein decided to suspend the expansion of the landfill until a thorough investigation had taken place.

In this weeks lab, we performed a suitability analysis regarding the placement of a new landfill in the fictitious Gallin County, Montana. We analyzed five aspects of the land- soil drainage, distance to water sources, landcover types, distance to other landfills, and the slope- and combined the factors to find the ideal spot for a new landfill. The analysis can be done in such a way which weights all factors equally, or more weight can be placed on factors that are more important, such as slope and soil drainage in this case. It ultimately shows the best place for something to go considering all the inputted factors.

In Kettleman city, although the landfill is already built, a suitability analysis would be beneficial to see whether or not the proposed expansion should take place. If the landfill is contributing to the birth defects, which is being investigated, certainly an expansion of the site would not improve the circumstances. If a suitability analysis is performed with factors such as distance from residential areas and distance from water sources, among others- and a better, more suitable landfill location is found, then perhaps a new one in the location may open up and gradually begin to replace the existing one.

A reverse suitability analysis of sorts may also be helpful in seeking out the cause(s) of the pollution and birth defects. We can look at the site of where the landfill is currently located and examine the land features surrounding it. We can, for instance, see how far the plant is from streams and what the soil drainage quality of the area is in which it is located, and calculate the likelihood of some of the pollutants from the landfill getting into public drinking water sources.

As a waste company that was fined $2.1 million for operating unauthorized landfills and waste ponds, and as a landfill itself that was in 2003 considered to be an emitter of excessively high radiation, it would not be too hard to believe that the plant was now contributing to local birth defects. However, in these types of investigations one must remain relatively unbiased to collect the necessary facts. Suitability analysis is a useful tool in this regard, but like in the application of all GIS tools, we mustn't be too reliant on it. There is always the possibility for the data to be off, or a human or software error to be made. An on site field investigation should be made in addition to a GIS suitability analysis in order to holistically research the issue at hand.

Wednesday, February 2, 2011

Challenge Quiz 1


This map shows how much space would be taken up in Los Angeles County if the new buffer zones were to go into effect, which would put out a lot of the population. As we can see with the 1000 foot buffers of the parks, schools, and libraries in LA, they take up a lot of area, and the most area also coincides with the places with the highest population density. If the buffer zones go into effect many will lose their access to medication and other dispensaries would become overcrowded. Certainly a compromise can be drawn between the safety of children and accessibility of medication for marijuana users, by perhaps proposing a 500 foot buffer as opposed to 1000. This would cut the buffer area in half while still providing safe environments for the children.

Wednesday, January 26, 2011

Week 3- Geocoding


For my geocoding lab, I decided to map out all of the In-N-Out Burger locations throughout Los Angeles county and show where they are in relation to the county's airports. I put two buffer rings around the airports, each representing two miles, so the buffers ultimately cover four miles from any airport in the county. I made it a point to clearly distinguish which In-N-Out Burgers fell within these buffer zones. I demarcated the three major commercial airports in the county, probably the three most relevant for this map, by labeling them with their three letter IATA (International Air Transport Association) codes: Los Angeles International Airport (LAX), Burbank's Bob Hope Airport (BUR), and Long Beach Municipal Airport (LGB).

I made this map because to many people, In-N-Out is a quintessential part of Southern California culture (although recently it has branched out to Northern California, Arizona, and Utah). One may argue that this food staple could be comparable to San Francisco's clam chowder and bread bowls. Many locals believe In-N-Out is a must have, especially for those who are visiting from out of town. Arriving into the Los Angeles area after a long flight, people may want to get something to eat right after they get off the plane. They may specifically want In-N-Out or decide they just want something to eat, as long as it's nearby and they can get it quickly. This map is for those people, who arrive at an airport and wonder if there's an In-N-Out nearby.

I went online and Google searched for address databases for various sorts of places but the closest thing I really found was on the in-n-out website, which was a printable chart listing the address of all their locations grouped by region, complete with concise driving directions to each one. For the reasons of easily accessible address data and an optimum amount of locations for the project within my target area, I chose to geocode In-N-Out. By means of of copying and pasting, I put all the addresses into an excel chart and I used an online zip code locator to find the zip codes for the addresses, which was necessary for geocoding using the address locator I was using on ArcMap.

It was very convenient having all the addresses being matched to their locations of the streets layer, and the ones that were initially unmatched I was quickly able to resolve. Had I had to place each point on the map myself, chances are that it would have taken much more time and the map would have less integrity. The buffer and select by location tools proved to be very useful in the creation of this map as they allowed me to easily where features were in relation to other locations. I can see geocoding coming in handy for the vast majority of GIS projects, and now that I have been introduced to it and have been applying it myself, I can see how important it is.

Data:

Addresses Zip codes
1210 N. Atlantic Blvd. 91801
420 N. Santa Anita Ave. 91006
324 S. Azusa Ave. 91702
13850 Francisquito Ave. 91706
761 N. First St. 91502
1371 N. Grand Ave. 91722
13425 Washington Blvd. 90291
21133 Golden Springs Rd. 91789
8767 Firestone Blvd. 90241
310 N. Harvey Dr. 91206
119 S. Brand Ave. 91204
1261 S. Lone Hill Ave. 91773
14620 E. Gale Ave. 91745
7009 Sunset Blvd. 90028
6000 Pacific Blvd. 90255
17849 E. Colima Rd. 91748
21620 Valley Blvd. 91789
3411 W. Century Blvd. 90303
14341 Firestone Blvd. 90638
15259 E. Amar Rd. 91744
2098 Foothill Blvd. 91750
5820 Bellflower Blvd. 90713
2021 West Avenue 93536
6391 E. Pacific Coast Hwy. 90803
4600 Los Coyotes Diagonal 90815
7691 Carson St. 90808
9245 W. Venice Blvd. 90034
25220 N. The Old Rd. 91321
5864 Lankershim Blvd. 91601
9858 Balboa Blvd. 91325
8830 Tampa Ave. 91324
13651 Roscoe Blvd. 91402
2114 E. Foothill Blvd. 91107
9070 Whittier Blvd. 90660
2505 Garey Ave. 91766
1851 Indian Hill Blvd. 91767
19901 Rinaldi St. 91326
3801 Inglewood Ave. 90278
4242 N. Rosemead Blvd. 91770
11455 Laurel Canyon Blvd. 91340
28368 Sand Canyon Rd. 91387
26401 Bouquet Canyon Rd. 91350
10525 Carmenita Rd. 90670
4444 Van Nuys Blvd. 91403
3640 Cahuenga Blvd. 90068
10601 E. Lower Azusa Rd. 91780
24445 Crenshaw Blvd. 90505
730 W. Carson St. 90502
6225 Foothill Blvd. 91042
7220 N. Balboa Blvd. 91406
2940 E. Garvey Ave. 91791
15610 San Bernardino Rd. 91722
9149 S. Sepulveda Blvd. 90045
922 Gayley Ave. 90024
19920 Ventura Blvd. 91364