Final Project: Suitability Analysis for Wind Energy Generators in Southern California

BACKGROUND
In more recent history, Americans have become more aware of the costs and consequences of using traditional coal and fossil fuels as energy sources. Now equipped with this knowledge and improved technology, the U.S. has increasingly turned to renewable energy sources such as wind energy. In the 1990s the State of California had long been the leader in generating wind energy, having the most windmill farms in the U.S. (CA Energy). Although Texas now generates the most wind energy in the United States, California's growing population demands increased energy that is renewable and sustainable. The State's varied terrain and land cover allows for additional sites for wind farms.

Inspired by the windmills along the Tehachapi Pass, I focused my research suitable locations within Southern California. This region of the state includes bustling metropolitan areas and burgeoning suburbs which have experienced increased immigration from other areas in recent years. Another point of interest in Southern California is that, compared to Central and Northern California's vegetation and agricultural land, the heavily urbanized areas in Southern California pose a challenge to finding space for wind farms. However, a large part of Southern California's land cover includes desert and grass which is more suitable for potential windmill sites.

METHOD
Before starting my project, I was posed with the question of defining which portion of the state constitutes "Southern California" versus Central or Northern California. Since these separate regions are culturally distinct from one another, residents of each region are particular (and even proud) about where they live. According to DeLio and Smith, Southern California includes, Santa Barbara, Ventura, Los Angeles, and San Bernardino Counties and southward, which are the counties included in my study.

Prior to my research, I considered several factors that would constitute a suitable location for a windmill farm, including: 1) landcover types, 2) elevation or slope of the area, 3) distance from urbanized areas (to provide energy for those communities), 4) impacts on bird migration paths, 5) regional wind speed averages, and 6) distance from already-established wind farms. While bird migration paths are a very important factor in siting for wind farms, I decided to exclude this factor, considering the time and resource limitations of this project.

In order to produce a map of the slope of Southern California Counties, I obtained Digital Elevation Models (DEMs) from the USGS Seamless Server. With ArcGIS I projected the rasters of the DEMs and then produced slope by percent rise. According to research by the State of California Energy Commission, higher elevations tend to yield faster winds, making areas of higher slope more suitable for windmill locations.
To convey the various types of land use and landcover throughout Southern California counties, I obtained data from the California Fire and Resource Assessment Program (FRAP). This data included the following categories: Agriculture, Herbaceous, Desert, Barren/Other, Shrub, Hardwood, Wetland, Conifer, Water, and Urban. I reclassfied the categories as follows (highest being the most suitable land for potential wind farms):

8: Barren/Other
7: Desert
6: Herbaceous
5: Shrub
4: Hardwood
3: Conifer
2: Wetland
1: Agriculture
0: Water
NoData: Urban


For distance in relation to urbanized areas, I obtained shapefiles of urbanized areas from the UCLA GIS Mapshare. From this data, I created 5-mile buffers from the boundaries of urbanized areas for two reasons: 1) to keep a reasonable distance from metropolitan areas, but 2) sometimes being closer may help make generating wind power more efficient.

To visualize the current locations of existing windmills, I geocoded the approximate addresses to avoid building turbines too closely to existing wind farms.

Lastly, from the State of California Department of Energy, I obtained data of annual average wind power in California (in miles per hour as well as meters per second). Seeing how this data spatially shows that some of the windiest on-shore areas in California are in Southern California counties.


Using the raster calculator, I reclassfied the layers of each suitability factor:

[ Land Use Type ] + [ Average Wind ] + [ Slope ] + [ Distance Urban Buffer ] + [ Distance from Existing Wind Farms ]


RESULTS
Reclassifying land use data revealed that a large portion of Southern California counties are desert, largely in San Bernardino and Riverside Counties. Naturally it makes sense that most existing California wind farms are located in these counties. Overlaying the layer of existing wind mills with the urbanized area layer revealed that windmills are currently fairly close to cities. Tjhis might suggest that the factor of distance from urbanized areas might not be as problematic as one might expect, depending on slope and other factors. In addition, while increased slope allows for faster winds, it may be more difficult to build turbines on high slopes.

After considering all five factors and running the raster calculation, it appears that the most suitable areas would exist in the eastern region of the San Bernardino Mountains. In this area there are higher wind speeds and higher slopes which are farther from heavily urbanized areas. Another suitable location appears to be in north San Bernardino County, just south of Kern County. While the Tehachapi and Mojave Wild Farms are situated in Kern County, just outside what is considered “Southern California,” these two locations are close enough to the southern counties to provide efficient wind energy.

CONCLUSION
Considering the factors of suitability included in this study, I think that a suitability analysis using this data would be helpful in locating potential windmill locations for Southern California. If it were not for time or resource limitations, I would certainly include analysis of avian migration information as it is extremely important when considering wind mill sites to minimize the incidences of avian deaths in the process of implementing renewable energy technologies.

Michael DiLeo, Eleanor Smith, Two Californias: The Truth about the Split-state Movement, Island Press, Covelo, California, 1983. pg. 9-30.

Pasqualetti M. J., Society and Natural Resources, Volume 14, Number 8, 1 September 2001 , pp. 689-699(11). Routledge, part of the Taylor & Francis Group.

Week 5 Lab: Interpolation

After reviewing the year-to-date data of precipitation levels in Los Angeles County, it appears that more recently there is less rainfall collecting in west Los Angeles County than during a usual season. The highest precipitation occurs in the eastern part of the County, particularly near the Angeles National Forest. While both the Kriging and Spline techniques of interpolation convey this phenomenon, the Kriging interpolation seems to present the precipitation data more generally, whereas Splining appears to capture the data points in between the gauging stations more. The splining technique also shows more variance in values between the gauging stations, which helps show greater difference when comparing the year-to-date and season normal data (last image). For these reasons, and since splining is best for representing smoothly-varying surfaces, I believe splining is the best technique for presenting rainfall information.

Quiz #2: Overview of ArcGIS

1. Rank order the ten most populous countries of the world. [6 points]
China
India
U.S.
Indonesia
Russia
Brazil
Pakistan
Japan
Bangladesh
Nigeria

2. How many rivers does the Amazon river system consist of? [6 points]
There are 15 rivers within the Amazon River system including:
Amazon
Guapore
Japura
Madeira
Madre de Dios
Purus
Putamayo
Rio Branco
Rio Juruena
Rio Maranon
Rio Negro
Rio Teles Pires
Tapajos
Ucayali
Xingu

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. [8 points]
There are 52 out of 2533 cities within 500km of the rivers.

4. To the nearest 100,000 what is the total population of countries within 300 kilometers of Iran (not including Iran)? [8 points]
When I used Select by location and pressed “OK” I reached the following results: Within 300 km of Iran, there are 48 countries. The total population of all of these cities, excluding Iran totals: 3,356,737,001. Rounded to the nearest 100,000, the population of these 48 nations is 3,356,700,000.

However, when I selected “Apply” rather than “OK,” ArcGIS gave me the total of 452,297,220 as the sum of these countries’ populations. Rounded, the population would total 452,300,000.

5. Identify the most and least populous countries of the landlocked countries of the world. [6 points]
The most populous landlocked country is Ethiopia with a population of 53,142,970. The least populous landlocked country is Vatican City, with a population of 860.

6. Identify all countries within 300 kilometers of Veszprem, Hungary (not including Hungary). [10 points]
Excluding Hungary, there are 9 countries within 300 km of Veszprem, Hungary. These countries include:
Poland
Czech Republic
Slovakia
Austria
Slovenia
Romania
Croatia
Bosnia & Herzegovina
Yugoslavia

7. What countries border Chad? [8 points]
Niger
Libya
Sudan
Central African Republic
Cameroon
Nigeria

8. Rank order of the five countries that have the most cities based upon the data. And what is the city number for each? [10 points]
To find the countries with the highest number of cities I used: Analysis Tools > Statistics > Frequency. I input the cities layer and CNTRY_NAME as the target layer. From this, the countries with the highest number cities resulted:

Russia – 97
United States – 93
Thailand – 72
Turkey – 67
Cote d’Ivory & Poland (tie) – 50

Part II [38 points + 5 bonus points]

9. Approximate the total length (km) of all river portions /segments flowing in the country of Sudan? [8 points]
The approximate total length of all river portions in the country of Sudan is about 6,500 km.

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? [10 points]
In order to find the top five countries with the most lakes in terms of number, I utilized the Frequency Tool as I did to find the highest number of cities by country. The five countries with most number of lakes are as follows:

Russia, 1516 lakes
Canada, 1340 lakes
United States, 743 lakes
China, 219 lakes
Sweden, 168 lakes

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? [10 points]
Canada: 443,517 sq. km.
United States: 196,849 sq.km.
Russia: 138,251 sq. km.
Kazakhstan: 70,900 sq. km
United Republic of Tanzania: 53,530 sq. km

In order to find this, I used the Geoprocessing Dissolve tool and added a field (in sq. km). I input the Lakes layer and dissolved the Lakes and Country Name columns, and ordered them by area (in km^2). The tool dissolved the Lakes and Country Name tables and ranked the countries by highest area of lakes (in sq.km.).

12. Produce the following map: a world country map of lake area per capita (area of lake surface per person). [15 points]

Week 4 Lab: Fire Analysis

Following the 2009 wildlife in La Canada Flintridge, California fire (also considered the “Station Fire”), authorities and local governments in Los Angeles County assessed the factors that led to such a hazardous and destructive fire. This assessment revealed that some of the major factors which attributed to the Station Fire were 1) the area’s increased or high slope in comparison to other less sloped areas and 2) the area’s types of land cover. Beginning my process of assessing slope and land cover for the Station Fire, I downloaded data from the California Department of Forestry and Fire Protection. I included data such as the widest extent of the station fire (perimeter), a digital elevation model (DEM) of Los Angeles County, and data featuring land cover types for Los Angeles.

From the Los Angeles County DEM, I projected the raster to obtain the slope values in percent rise to reveal that this portion of the Angeles National Forest had the most sloped area compared to other areas of Los Angeles County. When reclassifying the land cover data types, I assigned shrubs to have the highest fire risk and water having the least. Displaying the land cover types in Los Angeles County showed that nearly all of the land within the Station Fire perimeter was covered in shrub or hardwoods, which were the vegetation types with highest risk of burning. After preparing the slope and land cover maps, I combined the data and overlaid them to calculate the combined risk attributed by slope and land cover. As shown in the inset of LA County, the area of Angeles National Forest had the greatest combined fire risk, and the Station Fire perimeter is located entirely within this high risk area.

During the preparation of this fire assessment, some of the challenges I faced were largely based on software. For instance, I had to calculate slope/vegetation combined risk multiple times because ArcGIS wouldn’t allow the file name to be saved because it included too many characters. The raster calculator had a pop-up to indicate the software wouldn’t save the file, but the message it displayed didn’t make sense. Aside from such issues, there weren’t too many other challenges, and the data I analyzed revealed that these risk factors very obviously led to such a high risk for the Station Fire to occur. Studying fire ecology typically shows that over time, it is natural for areas of forest and shrub to have small fires in order to maintain the ecosystems of these areas. Some hardwoods require fire to open and release their seeds and continue growth within an area, thus allowing for a cycle of burn and regrowth. Suppressing forest fires will tend to lead to large fires that will burn long, such as the case with the 2009 Station Fire.

Lab Week 3: Raster Suitability Analysis

Final Project Topic Proposal - Wind Energy in Southern California

As one of the most highly-populated states, California’s population is always in need renewable, sustainable energy sources. For my final project I propose to find the most suitable potential locations for a windmill farm in Southern California. The project would begin by geocoding current locations of windmills in the region and assessing the amount of energy that is produced and where this energy is used. The preliminary research will include studies of known bird migration studies to mitigate the impact of avian deaths caused by windmills, as well as records of areas within the state that are sources of strongest wind paths.

I would be able to find data from the State of California Energy Commission and population data from the U.S. Census Bureau to analyze which populated regions would be most benefit from increased wind energy. This project will also consider elevation and slope to find the most suitable areas.

Week 3: Quiz #1 - Medical Marijuana Dispensaries

In regards to the recent Los Angeles City Council policy to close all medicinal marijuana dispensaries within 1000 meters of Los Angeles schools, I take the position of support for this ruling. Through GIS analysis of schools, parks, and libraries in the City of Los Angeles, the spatial evidence that supports my argument most is the case for the neighborhoods of Southeast Los Angeles and Hollywood. After mapping the locations of Los Angeles County libraries, it is apparent that the highest concentration of libraries exists in Southeast Los Angeles, near cities including Vernon, Huntington Park, Cudahy. Despite the City’s recent ruling to close dispensaries within 1000-meters of schools and libraries, there still appear to be many dispensaries within miles of schools (see: Map of Southeast Los Angeles).

Another primary point of concern is the LA neighborhood Hollywood. Hollywood, world-famously known for its entertainment district and destination for recreation and the arts, some would argue that it is an obvious location for many dispensaries. Walking the streets of Hollywood, however, most people would not readily notice that there are many schools within this neighborhood behind all of the theatres, restaurants, and centers of entertainment. As seen on the map of Hollywood showing locations of schools and their buffered distances from dispensaries, it is apparent that there are several dispensaries within 1000-meters or just outside the proposed buffer zones.

Week 2 Lab: Geocoding

For this week's assignment, I chose the topic of ecological reserves and protected areas within Orange and Los Angeles Counties. While researching and collecting data, I found that while there are many small parks within cities, there are far fewer stretches of land that are protected for their plant or wildlife biodiversity. In urbanized areas many people are familiar with local neighborhood parks, which residents primarily use for recreational use. While areas near ecological reserves and wildlife sanctuaries are often open to recreation including hiking and equestrian trails and camping, state and local governments reserve these areas to protect native plant and wildlife.

Aside from outdoor adventurers, researchers, or land stewards who show interest in the biodiversity of these areas, within urbanized areas, the general public tends to be less aware of these reserves and their conditions. In this case, geocoding is applicable to assigning relative addresses to these areas to inform LA and Orange County residents that they live within a biodiversity hotspot and of the often rare species that exist in these places.

Viewing Los Angeles and Orange County ecological reserves in relation to the major highways in these counties reveals that contrary to what one would expect, most of these areas exist within or within a one-mile radius of urbanized areas, and usually immediately adjacent to major roads. Studies of biogeography show that the boundaries of natural areas are often compromised when nearby urbanized areas or infrastructure. However, perhaps the close proximity of these areas to urbanized areas and roads can assist locals in attempting to protect and study wildlife reserves. In addition, while gathering data I also found that many parks and reserves do not have officially addresses. Since, over time, the borders of natural areas continuously change, geocoding may be useful in tracking the where certain parts of these natural areas exist in the past, present, and future.

The ecological reserves within Los Angeles and Orange Counties are protected by the California State Department of Fish & Game, and Los Angeles and Orange County both report that these areas are often habitat to rare wildlife species as well as wild plant species. In addition to rare wildlife, many of these natural areas have very unique geology, suitable for studying geomorphological change. Protection and awareness of these areas and their unique species would allow for further study of Southern California wildlife and geomorphology.