Ecology of Human Habitat and Urban Form

CMP3200 ASSIGNMENT 4 Spring 2018

The Ecology of Human Habitat and Urban Form

In this packet:

· Assignment description

· Example

· Data tables (5 tables)

· Also see paper by Wheeler in a separate pdf

The field of urban design looks at the form of urban areas, that is, the patterns and shapes of how cities occupy landscapes. If you think about the straight grid pattern of many modern cities, versus the narrow, twisting streets and alleys of ancient cities, versus the “loops and lollipops” of American suburbia, you are thinking about different urban forms. There is also a vertical element to urban form, again thinking about high rise buildings versus low-density single-family residential neighborhoods. Intuitively, we can imagine that these diverse urban forms have implications for ecological systems, as well as for how we, humans, experience these varied “human habitats”.

In this assignment, you’ll use a typology (a classification system) of urban forms and relate these to data on ecological variables that may affect the long-term sustainability of a city and its livability as human habitat. You’ll start by coming up with a hypothesis about how these variables might relate to one another, then organizing the data you need to test that hypothesis.

MATERIALS:

· This assignment document that includes instructions, a sample assignment, and data tables.

· A pdf of a paper by Stephen Wheeler describing the urban form typology (on Canvas).

· An Excel workbook with the same data tables that are in this assignment document. Each table is in a SEPARATE SHEET – look for the tabs along the bottom to switch between sheets.

DATA GUIDE:

Table 1 is a list of the urban form types, with descriptions. Further details and illustrations are available in Wheeler (2015) available here.

Table 2 is a breakdown for 5 American cities of the total area (in square km) represented by each urban form type. Note that not every city has every type.

Tables 3-5 contain calculations of variables for each urban form type for each city as follows:

Table 3 shows average surface temperature in degrees C on a typical August day.

Table 4 shows average impervious surface percentage. Impervious surface is the areas of the land surface that are covered in materials that do not let water percolate through to the underlying soil, including bare rock, pavement and rooftops. The opposite of impervious surface is pervious surface, including bare soils and vegetated areas.

Table 5 shows “vegetation” as calculated by NDVI (Normalized Difference Vegetation Index). NDVI is based on satellite imagery of the land surface and is based on the difference in how different wavelengths of light (especially green) interact with live vegetation versus other surfaces. Thus it’s simply a measure of live vegetation “green-ness”. (Surfaces that are just green-colored won’t be included). The larger the number, the more green and vegetated the surface is. Note that you might expect green-ness of an area to be negatively correlated with impervious surface. ..

Some examples of observations we can make looking at these tables:

From Table 3 we can see that on a typical August day in Sacramento, the average surface temperature in a “Loops and Lollipops” area is 38.25 degrees C, while in a “Malls and Boxes” area the average surface temperature is 41.67.

From Table 4, we can see that the average impervious surface percent for “Commercial Strip” ranges from a low of 11% impervious fraction in Sacramento to a high of 71% impervious fraction in Portland.

From Table 5, we can see that vegetation is very low in airports (unsurprisingly) and quite high on hillsides and county roads.

PROCEDURE:

1.Spend some time with the data tables, making sure you understand what they represent and how they might fit together. Read Wheeler (2015) to understand the urban typology classifications. You have the SAME data tables in hard copy here and in an Excel file on Canvas.

2.Come up with a research question and hypothesis that you can test using the data available. You will probably want to cut and paste some rows or columns to create a customized table that meets your needs. In the example below, my custom table didn’t use any of the existing rows or columns and I just found it easiest to enter the values I wanted by hand from the printed tables.

3.Test your hypothesis, either by running a simple statistical analysis like a t-test or regression (see Data Analysis under the Data tab in Excel) OR just by graphing the data in some way so that you can visualize the relationship between the variables.

To hand in (see example):

1. A research question, and a hypothesis, with justification. That is, what do you want to find out, and what do you think the answer is and why?

2. A rearranged, custom data table you used to test your hypothesis.

3. Some sort of test of your hypothesis – this can include a simple statistical test like a t-test or simple regression, or maybe just creating a chart or graph that allows you to visually compare the variables.

4. A short discussion of your findings and conclusions.

NOTES:

This assignment involves using Excel spreadsheets. If you aren’t proficient in Excel, Kate or Sarah will be happy to sit down with you and show you the basics. If you don’t have Excel, it is available on pretty much any University computer – try the library. You are NOT expected to use statistics unless you are familiar and comfortable with the analysis you want to do. It is best in this case to do something simpler but to do it well.

You are not expected to use all of the data tables! They are there for you to pick and choose from. In my example I did not use the Land Area data, but you can if you want to. That data table basically provides the option of calculating some observation by the area of its landscape type for each city, so that you can know how much of a city’s area is represented by that type. You should use data from at least two of the tables though, or you won’t have a very interesting research question.

GRADING:

You will be graded on:

· your ability to formulate a clear research question and hypothesis that show you understand the data, and can think about how these variables relate (2)

· creating a data table that presents the data in a way that is appropriate to your hypothesis (1)

· presentation of a graph or statistical analysis that is accurate and appropriate to the research question (1)

· being able to tie your findings back to your hypothesis in your discussion – What have you learned here? (1)

EXAMPLE

1. Research Question: Are urban form types with higher vegetation values cooler across all cities?

Hypothesis: I predict that in the month of August, the greater the vegetation component of the landscape, the cooler the surface temperature because vegetation provides both shade and the cooling effect of transpiration by plants.

2. I selected 4 urban form types that were present in all 5 cities and had high and low values of NDVI. Since the variables I want to compare are Vegetation and Temperature, I just created a column for each, therefore the data are displayed by city but they are not sorted by city in my analysis or in my graph.

City	Urban Form	Vegetation NDVI	Temperature degC
Sacramento	Commercial Strip	0.2	40.74
	Workplace Boxes	0.28	41.32
	Loops and Lollipops	0.36	38.25
	Upscale Enclave	0.31	37.16
Portland	Commercial Strip	0.25	41.6
	Workplace Boxes	0.26	42.76
	Loops and Lollipops	0.49	38.29
	Upscale Enclave	0.58	36.17
Las Vegas	Commercial Strip	0.17	51.97
	Workplace Boxes	0.22	52.32
	Loops and Lollipops	0.35	49.86
	Upscale Enclave	0.38	48.2
Atlanta	Commercial Strip	0.37	43.14
	Workplace Boxes	0.38	45.06
	Loops and Lollipops	0.54	37.94
	Upscale Enclave	0.6	40
Boston	Commercial Strip	0.29	35.37
	Workplace Boxes	0.45	36.06
	Loops and Lollipops	0.51	32.59
	Upscale Enclave	0.64	30.79

3. I created a scatter plot of my two variables to see if there appears to be a relationship between the amount of vegetation and temperature. Because my hypothesis predicts that the amount of vegetation is affecting the temperature, I plotted vegetation on the x axis and temperature on the y axis. When I plotted them, there did appear to be a negative relationship, that is, the higher the vegetation index, the lower the temperature. In order to check it further, I ran a regression on the two variables, again with vegetation as the independent variable (x) and temperature as the dependent (y). The regression analysis showed an R2 of 0.377 with a p-value of 0.004.

4. The graph and the regression analysis support my hypothesis that temperature is negatively related to the amount of live vegetation on a summer day. My R2 value from the regression indicates that vegetation explains about 38% of the variability in the temperature data and since the p-value is small, this means that there is a low probability that this result occurred by chance. I noticed as I worked with the data that the higher values for NDVI varied a lot between cities. Las Vegas had a maximum green-ness of .38, while Boston’s maximum was .64, and the two cities had very different temperature ranges as well. This isn’t surprising given the very different climate in these two cities, and I expect that this difference is behind much of the additional variation in the data that my regression couldn’t explain. On the whole, my analysis supported my own experiences, namely that on a hot August day in pretty much any city, I would find it to be cooler in a leafy green suburban residential neighborhood than in a strip mall parking lot.

TABLE 2: Land area covered by built landscape types (square km)
	Sacramento	Portland	Las Vegas	Atlanta	Boston
Airport	47	18	23	32	25
Apartment Blocks Campus	0	0	0	0	8
Campus	55	38	17	64	31
Civic	47	1	2	3	2
Commercial Strip	24	21	25	160	48
County Roads	1	0	0	6	72
Degenerate Grid	106	182	66	107	429
Garden Apartments	43	26	49	122	30
Garden Suburb	14	4	0	1	84
Heavy Industry	32	74	12	41	17
Hillside	4	8	0	0	0
Incremental/Mixed	37	27	94	68	148
Land of the Dead	2	4	1	7	22
Long Blocks	0	0	0	8	0
Loops and Lollipops	449	305	352	2469	601
Malls & Boxes	38	23	48	65	64
New Urbanism	4	7	0	1	0
Organic	0	0	0	0	0
Quasi Grid	6	1	0	25	121
Rectangular Block Grid	14	72	17	10	120
Rural Sprawl	954	291	13	943	1024
Superblocks	0	0	0	0	0
Trailer Park	13	8	9	4	6
Upscale Enclave	135	16	46	32	50
Urban Grid	27	35	1	13	1
Workplace Boxes	102	6	51	299	144
Total	2156	1194	824	4479	3047
Approx. population, millions	1.9	2	2.2	5.5	4.5
Density (pers/sq. km)	881	1675	2669	1228	1477

TABLE 3

TABLE 4

TABLE 5

Temperature degC

0.2 0.28000000000000003 0.36 0.31 0.25 0.26 0.49 0.57999999999999996 0.17 0.22 0.35 0.38 0.37 0.38 0.54 0.6 0.28999999999999998 0.45 0.51 0.64 40.74 41.32 38.25 37.159999999999997 41.6 42.76 38.29 36.17 51.97 52.32 49.86 48.2 43.14 45.06 37.94 40 35.369999999999997 36.06 32.590000000000003 30.79