Tuesday, September 13, 2011

Final Project, Does Income Influence the Prevalence of Obesity?


I. Introduction

            Obesity has become an increasing epidemic in the United States in recent years. About 60 million adults, or 30% of the adult population, are now obese.  And since 1980, overweight rates have doubled among children and tripled among adolescents –increasing the number of years they are exposed to the health risks of obesity.  In fact, in Los Angeles alone, 14% of students were obese. Also, based on a large national study, body mass index (or BMI, an indicator of excess body fat) was higher every year between 1986 and 2002 among adults in the lowest income group and the lowest education group. Given these facts, the question we ask ourselves is if income has any sort of influence on the prevalence of obesity, particularly in Los Angeles County. Two cities were chosen to be the focal areas due to their significant income difference- Beverly Hills and Downtown Los Angeles and McDonalds was chosen fast food restaurant because it is the most popular and most prevalent fast food restaurant in America. With the question in mind, three maps were created, one showing the obesity rates in the Los Angeles County health districts, another showing the median household income rates, and the last showing McDonalds locations in two specific areas, Beverly Hills and Downtown Los Angeles. Lastly, a final map was created integrating all the parameters together.

II. Methods

            To create these maps, my group members and I first read about the recent concern on obesity in Los Angeles. We found a PDF on an article written by the County of Los Angeles Department of Health Services, called “Obesity on the Rise” from July 2003. The article provided many charts and statistics about the increase of obesity rates based on the department’s own research. The article also exhibited a map of Los Angeles County indicating the prevalence of obesity in the different health districts. We first wanted to extract that exact map, so we researched for that specific shapefile, but had no luck. But we found the Los Angeles County health districts shapefile so, using some GIS techniques, we added a field to the attribute table and added the percentages of each district to the map and changed the color ramp to indicate the rates. We also highlighted the specific cities we’re focusing on, Beverly Hills and Downtown Los Angeles, using a shapefile of Los Angeles County’s zip codes and determined which zip codes belong to their respected cities and selecting those specific zip codes using the attribute table. For the next map, we first gathered all the addresses of the closest McDonalds locations to the focal cities and put them on an Excel spreadsheet. With experience from Lab 2, we used geocoding to input those locations to create another map. create the last map, we had to research the income levels in Los Angeles County. We found a website that listed the median household income averages in each zip code. So with this information, we made another Excel spreadsheet and exported the data to ArcMap and changed the color ramp to classify the income rates based on their zip codes.


III. Results
           
            Looking at the first map, Prevalence of Obesity in Los Angeles, you can see that there is a significant difference in obesity rates in Los Angeles, particularly comparing the Westside to the Eastside. The Eastside has the highest obesity rate, with < 25% of the population being overweight. Looking at the two focal regions, Beverly Hills, outlined in blue, is located in the 16 - 20% obesity region, whereas Downtown Los Angeles, outlined in green, is located in the 21 - 24% region and the < 25% region. In the next map, after geocoding the McDonalds locations, we can see where the restaurants lie in within the two focal regions. As indicated, there are no McDonalds locations within the Beverly Hills region. And there is a significant difference in numbers of restaurants in Beverly Hills compared to Downtown Los Angeles. In Downtown Los Angeles, there are about twenty McDonalds locations where there are none in Beverly Hills, but there are seven locations nearby the Beverly Hills zip codes. And for the last map, we can see the classification of income rates based on the colors, comparing the colors in Beverly Hills and the Downtown Los Angeles region. In Beverly Hills, the median household income was over $75000 but in the Downtown Los Angeles region, the median household income ranges from $0 to about $25000.

IV. Conclusion/Discussion
           
            By looking at the outputs, we can conclude that there is a correlation between income and obesity levels in Los Angeles County. We have learned that there are no McDonalds restaurants located in the 91210 zip code and the median household income is above $75000. We can make the assumption that those with higher income don’t eat cheap fast food but still eat unhealthily because there is still a significant obesity level in the area. Those with lower income, due to the decreasing value of the economy and higher levels of unemployment, eat what they can afford and fast food is cheap and most accessible, especially in the Downtown Los Angeles region. We didn’t include other popular fast food chains but using fastfoodmaps.com, we saw that are so many more restaurants in the Downtown Los Angeles region, such as Burger King, Taco Bell, KFC, and Carl’s Jr. In Beverly Hills, the closest fast food restaurant is a Taco Bell, but its location still doesn’t belong in the 91210 zip code. It was startling to see that there were no popular fast food restaurants in Beverly Hills at all. It may explain why the obesity rate in the area isn’t as high as it is in other areas such as Downtown Los Angeles. I guess you can say that the privileged in Beverly Hills have better food options compared to those living in Downtown Los Angeles. That statement can be true within the UCLA demographic alone. For example, I have friends who try to eat organic and locally grown produce but shopping at Whole Foods can really put a deep dent into an unemployed college student’s wallet, so many of us end up eating cheap fast food in Westwood to relieve our wallets and satisfy our stomachs. Its also easier and faster to go to a drive-through than have to cook your food then clean it up afterwards. It is about convenience for us Angelenos. Because we live such crazy and hectic lives, we eat poorly, but because it is cheap and fast, we continue to consume without thinking about the consequences and how it affects our health. I really enjoyed making these maps using ArcGis with my group because we definitely learned a lot and got to use our ArcMap skills we have accumulated from the past couple of weeks. We refined our ArcMap skills by making the three maps and learned valuable information. It definitely got me to think about our growing population in Los Angeles and how fast food has affected our health and will continue to affect us. There is no way to completely get rid of all the fast food locations here but we can always make the healthier choice by reducing our fast food intake. 


Data Sources-



Los Angeles County and Los Angeles County Zip Codes shapefile


Median Household Income
http://zipatlas.com/us/ca/los-angeles/zip-code-comparison/median-household-income.htm


McDonalds locations
http://www.fastfoodmaps.com/
http://www.mcdonalds.com/


Prevalence of Obesity
County of LA, Department of Health Services, Public Health
http://publichealth.lacounty.gov/ha/reports/habriefs/lahealth073003_obes.pdf


Group Members- Ellyse Briones, Eric Ching, and Lisa Tse

Thursday, September 8, 2011

Lab 5, Spatial Interpolation



         In this lab, we learned how to perform spatial interpolation to show the amount of rainfall in Los Angeles County. Using the same data, it is clear that the results are different depending on the used methods. I used the Inverse Distance Weight method for my first map set and the Kriging method for my second map set. Both map sets show rainfall throughout Los Angeles County from October 2010 till September 5, 2011. These maps show the normal, seasonal, and difference of rainfall in Los Angeles County. Analyzing the maps, you can see that precipitation levels the mid-eastern area in Los Angeles, particularly in the San Gabriel Valley, were the highest and in the northeast, in the Antelope Valley, were the least. This is because the Los Angeles Forest, having high mountain ranges, is located in the San Gabriel Valley, and in the Antelope Valley, it consists of deserts, proving the reason for those precipitation levels. Looking at the IDW map set, the Normal and Seasonal maps are very similar but with the Kriging method, none of the maps are similar. Comparing the two map sets, I think the IDW method better illustrates the rainfall data, simply because you can see the differences better than compared to the Kriging maps. Also, the IDW maps are easier to understand with the chosen colors. For example, comparing the Seasonal maps from the two methods, the Kriging output shows the colors in a distorted fashion, making it rather difficult to understand. Also, the Kriging maps seem to overlap the precipitation levels, further making it harder to understand and analyze.  Lastly, it is harder to predict rainfall levels using the Kriging map sets compared to the IDW maps. Using these methods are definitely beneficial to analyze rainfall but I think the IDW method is the better method because the data is illustrated comprehensively and clearly.  With the used data, the IDW method is effective to interpolate the points based on distance and is better suited for what we are trying to illustrate.


Wednesday, August 31, 2011

Quiz #2, GIS Challenge

Part I
1. Rank order the ten most populous countries of the world.
1. China
2. India
3. United States of America
4. Indonesia
5. Russia
6. Brazil
7. Pakistan
8. Japan
9. Bangladesh
10. Nigeria

I selected the country layer on ArcMap, right click to show attribute table. Scrolled to see the POP_CNTRY, right click to put in descending order.

2. How many rivers does the Amazon river system consist of?
15 rivers- Amazon, Guapore, Japura, Madeira, Madre de Dios, Purus, Putamayo, Rio Branco, Rio Juruena, Rio Maranon, Rio Negro, Rio Teles Pires, Tapajos, Ucayali, Xingu

I selected the country and rivers layer, right click to show the attribute table. Under system, I doubled clicked so it sorted according to the system.

3. How many cities are within 500km of the Amu Darya and Syr Darya rivers? Attach a screen shot of a table for these cities.
37 + 52= 89 cities

 
I clicked the attribute table of the river layer, selected the 2 rivers. Then I clicked Selection > Select by location > Selection Method is select features from cities and the source layer is rivers > Spatial Selection method is "features within a distance of source layer", check apply a search distance, typed 500 km, then clicked OK. The region would be shown, then clicked the cities layer and opened the attribute table.  

4. To the nearest 100,000 what is the total population of countries within 300 kilometers of Iran (not including Iran)?
452,300,000
I first select by attributes > layer: country  > name > "CNTRY_NAME" = 'Iran'. Then I
Select by location > Target layer: cntry02 > source: cntry02 > Spatial selection method within a distance is 300 km and clicked OK. Then I opened the attribute table from cntry02 layer with the selected feature > statistics on POP_CNTRY.


5. Identify the most and least populous countries of the landlocked countries of the world. 
 Most populous country- Vatican City
Least populous country- Ethiopia

With the country layer selected, I right-clicked to go to the country attribute table. Scrolled to see the Landlocked column, selected the Y of landlocked countries, then clicked Show selected records. Then I right-clicked on pop_cntry > Sort descending. Then I looked for the most and least populous countries.

6. Identify all countries within 300 kilometers of Veszprem, Hungary (not including Hungary).
Poland
Czech Republic
Slovakia
Austria
Slovenia
Hungary
Romania
Croatia
Bosnia & Herzegovina
Yugoslovia 


I clicked the Attributes table of country > selected Hungary> Select by location > "Select features from" > Target layer is cntry02 > Source layer is cntry02 > Spatial selection method is "Target layer within a distance of the source layer feature" > and clicked OK. Then I opened the Attribute table of the country layer > Show selected records, excluding Hungary.

7. What countries border Chad? 
Libya
Niger
Sudan
Nigeria
Central African Republic
Cameroon

I clicked the Attributes table of country > selected Chad > Select by location > "Select features from" > Target layer is cntry02 > Source layer is cntry02 > Spatial selection method is "Target layer features touch the boundary of the source layer feature" > and clicked OK. Then I opened the Attribute table of the country layer > Show selected records, excluding Chad.

8. Rank order of the five countries that have the most cities based upon the data. And what is the city number for each? 
1. Russia- 97
2. United States- 93
3. Thailand-72
4. Turkey-67
5. Cote d'Ivory and Poland (tied)- 50


I  first opened ArcToolbox > Analysis tool >Statistics > Frequency >Input table- cities> Output table > Frequency field- Country name > OK. A new table will show up, so I opened the table > right-clicked frequency > sort by descending.

Part II 

9. Approximate the total length (km) of all river portions /segments flowing in the country of Sudan.
The rivers flowing in the country of Sudan are the Nile, the White Nile, and Blue Nile, approximating 3920km.


2936km +361km + 623km = 3920 km


I first selected Sudan as the location, then selected the rivers in Sudan, the Nile, White Nile, and Blue Nile. Then using the measure tool, I clicked down, following each river through Sudan and putting down each length for each segment and added them together.

10. Rank order of the five countries that have the most lakes in terms of number. And what is the lake number for each of the five countries?
1. Russia- 1516
2. Canada- 1340
3. United States-743
4. China-219
5. Sweden-168

I  opened ArcToolbox > Analysis tool >Statistics > Frequency >Input table-lakes> Output table > Frequency field- Country name > OK. A new table will show up, so I opened the table > right-clicked frequency > sort by descending.
 
11. Rank order of the five countries that have the most lakes in terms of area. And what is the total lake area (square km) for each of the five countries?


12. Produce the following map: a world country map of lake area per capita (area of lake surface per person).