Land Use/Cover Data -- Student Worksheets

Student Worksheet 2.1
Activity 2.1 What's the problem anyway???

Let's begin then by assuming that we want to know whether there really is a relationship between human driving forces and land use/land cover change. What do we need to know and do in order to answer this question? How do we approach this question? First of all, remember that there are four major driving forces, and there are many types of land use and land cover. So what we want to answer turns out to be a really huge question! Take a moment to think about how you would approach the problem. You may discuss this with other students in your group.

A What we are trying to do is turn a broadly stated "problem" into a "researchable" question. As a first step this requires, as you probably discovered yourself in your discussions, that we need to cut the problem down to "bite-size." On a separate piece of paper, try to re-define the problem in which we are interested in a one sentence question. Feel free to limit this question to just one of the driving forces, one or few land uses, and to a reasonable time frame (what time frame is reasonable? why?). You need to make a decision as to what is most worth knowing, most important to you! (If you can=t decide right now, write down several questions that interest you). This step is called problem formulation.

When you are finished, exchange your sheet of paper with your neighbor, and think about what assumptions he or she must have made to write the question the way he/she did. For example, does the research question sound to you as if he or she was saying "population growth is the most forceful of all driving forces" or "natural land cover is always better than any human-altered land cover?" Give your intuition free reign! Take a few notes on a separate sheet of paper, and after a few minutes tell your neighbor what you found. Ask whether you are in the right ball park and discuss your assumptions with each other. If you can, imagine what the same kind of question might sound like under different assumptions.

B Now we have a clear question. But do we really? Say you chose to look at economic change over the past 50 years in the United States as the driving force, and you want to see if that is related to change in forestry. Your next question must be: how do you measure economic change or change in forestry? The terms in quotation marks are called variables, categories that in themselves are not constant but variable over time and/or space. (Other examples may be health, climate, wealth, ideology, etc.). Determining an actual measure that expresses or stands for the variable of interest is called operationalization. So in our example, a measure of economic change may be the change from one year to another in Gross Domestic Product (GDP), or export volume or $ of investment, etc. Change in forestry may be indicated by amount of land under all types of forest (in ha) or number of employees in the (US) Forest Service. Notice how many different measures for the same variable there can be! What difference does it make for the interpretation of your analysis which variables and measures you choose? Operationalize the variables of your research question, and choose those measures that make the most sense to you. Why did you choose the ones you did?
 

C Next, determine what kind of data you would need according to the measures you have chosen. Assume that you don't have them at hand. Where would you get them? In what form do you need the data? Do the data exist in that form, say in tables, on maps, in computerized databases? Do you need to do field work, measure something, acquire remote sensing imagery? This step of getting the data you need is called data acquisition. Note in the table on the next page the most important kinds of data, the form in which you would like them, and where you could acquire them.
 

D Now imagine that you have put in weeks of hard labor, long phone calls, spent quite a bit of money, and you finally have in front of you a pile of dusty files, maps, statistical yearbooks and more. How do you get from this pile of information to your data, not to mention the answer to your question? Notice the fine distinction between information and data. Not every bit of information you will find in this pile is actual data that you can use to answer your research question. Some of it may be outdated or may not apply to the time frame or geographic scale you chose; some of it may have nothing to do with your question. In case you want to aggregate data from different sources (a process to be done cautiously at any rate!), not all possible data may lend themselves to this compilation, and so on. The only way to find out what you can use and of what quality the data is, is through the often tedious process of data assessment.

We will not do a complete data assessment here (in Activity 3.3 we will come back to this). But let's focus on the difference between information and data. Use the hand-out provided by your instructor, and discuss with your neighbor what data you believe is pertinent to the research question that was posed along with the hand-out and what you believe is "just" information. Both of you should take notes on what you discuss.

E Since you did not actually collect information, we have no data to assess. But imagine you had done that, and found that most of it is of no use to you. Depending on why the information you gathered was useless to you, you might have to gather additional, better, or other data. You might even have found that your question is not answerable with the data that exist, and that you either have to generate your own data or reformulate the problem, and then go through the process all over again. For the purpose here, let's assume the "convenient" case of having enough and good-quality data for our investigation. The next logical step then is the data analysis. A large number of methods can be employed to analyze data, and the choice of methods depends both on the kind and quality of data and on the purpose of the analysis, i.e., the question you want answered.

Get together with your neighbor again and brainstorm about how to analyze data. For example, imagine you had the data needed for the example on the overhead that your instructor showed you: the population figures for the US for 1850 through 1990, and the numbers for wheat yield for the same years. What could you do with that? How could you use these data to answer the research question as stated? Or else, use your own research question and the list of data that you thought you needed to answer that question. How could you use those data to find your answer?

F Closely interlinked with, but for the purpose of clarity here separated from, the data analysis is the final step in answering our research question: interpretation of the results of our analysis. Principally, two outcomes are possible here: either we successfully were able to answer our question, or we were not and have to re-do our analysis, find more or different data, or start all over with a different question.

Since we didn't do an actual analysis, we can't interpret any specific results. But you can imagine that your research may either lead to a satisfactory answer, open up new ways to think about a problem, help you see ways to manage the problem, or it may allow you to see new connections between facts, and new relations between environmental and societal processes. All of this would be the result of your interpretation of the research results.

Student Worksheet 2.2

Activity 2.2 Getting Wired for Global Change Research

This "dry run" through the research process showed us at least two things. First, problem formulation is the most crucial step in any type of research; everything else hinges on that initial step! Your final interpretation inexorably depends on the way you asked the question and then how you operationalized the important variables, the data you sought and used after your critical assessment, and the analyses to which you subjected the data. Recall that your classmates probably stated the research question differently from you, and that their operationalization and analysis was just as reasonable as yours. But how did each one of you answer the initial problem? Even the most objective and reasonable research always contains elements of subjective judgement because of the choices we must make in the research process.

Second, it has become clear that the research process is neither a straight nor an easy road from a problem to a solution. Both of these points are important to keep in mind in assessing one's own and other people's research. For the flow chart below (Figure 4), recall each step of the research process, and fill in the blank boxes so that you end up with a logical sequence.

Student Worksheet 2.3

Activity 2.3 Naming It -- Counting It: How Terminology Matters

In this activity, we will try to figure out how to identify "good" data; in a manner of speaking, data that are worth gold, not just fool's gold. We will begin by looking at land use/land cover data over time and consider how data gathered by the same researcher or agency may differ simply because the definition of the measured variable changed somewhere along the way. Recall that Skole in his paper said (Skole 1994: 442), "data from the same source may vary considerably from year to year due to changes in methodology or terminology. [As Figure 5 ], time series derived from later editions of the U.N. Food and Agriculture Organization (FAO) Production Yearbooks, an important source for this kind of [data from] recent history, differ from the same time series derived from earlier editions." Let's check that out!

Go to your college or university library and find the FAO Production Yearbooks. Look up editions of 1970, 1980, and 1990 (or use the hand-outs provided by your instructor), and note the definitions used for arable land, permanent cropland, permanent pasture, and forest land. Take notes on what you find. Are there any differences in the definitions? If so, do you feel they matter?

Next, find any description for how the data on these categories were obtained, what they include and what not, to what time span they refer, and how the methodologies to aggregate or obtain the data in the first place differ from decade to decade? Take notes on all of this or make yourself copies of the relevant pages. Then meet in small groups (of three or four) and compare and discuss what you found in the Yearbooks. You should think about the following questions:

• What do these differences mean for the quality of your data set?
• Given these data, how confident are you about the actual change in land cover shown in this figure?
• If you were to cautiously interpret these time series, what could you say?

Student Worksheet 2.4

Activity 2.4 Reading Between the Points ...

• whether data are available for the time span that we are interested in,
• what their quality (completeness, consistency, comparability, etc.) is, and
• which of the data to use in case there are more data than we can feasibly include in our analysis.

A time series, as the term implies, is a sequence of chronologically ordered data values. For example, if you measured the outside temperature every day and ordered the measurements by date, you would have a time series of temperature data.

Interpolation is a method employed for estimating values in between sample points (Figure 6). Paraphrasing the common idiom, interpolation can be thought of as "reading between the points." Usually, you would do that by following the trend that the existing sample points suggest or by relying on a pattern that you expect to underlie their distribution. The following graph shows this:

Sampling is the process of selecting your objects for analysis from a potentially infinite number of objects with similar characteristics (called a population). Since not all members of a population can be incorporated in the analysis, especially when the population is very big, a specific (purposeful) or random (representative, bias-free) portion of it needs to be drawn to obtain clues about the entire population. Sampling schemes determine the way in which this portion is obtained. For example, one could conduct a survey of American farmers to obtain data on their land use practices, but because there are thousands of farmers (the population), we decide to survey only a subset of them (a sample of maybe 250) drawn randomly from addresses available from farm bureaus in different regions of the country (the sampling scheme).

What does this have to do with land use change? Read the little box of information below and then go on with this exercise.
 

Carbon is one of the most common chemical elements on earth. It can be found in the earth's atmosphere as CO and CO2 and as one of the most essential components of organic materials (e.g., in plant or animal tissue). When this material dies off and decays (or burns), the solid carbon compounds are released as gaseous carbon compounds to the atmosphere. The invisible, odorless gas CO2 is very significant in this respect because it has the capacity to let solar radiation into the atmosphere, but not back out. It "traps" heat like a greenhouse. Hence the term "greenhouse effect" -- a natural phenomenon that is enhanced by human-induced releases of CO2 and other radiatively active gases.

Since the change in carbon storage in vegetation (i.e., carbon flux) is predicted from the change in the area under cultivation through a mathematical formula (called a model), the three graphs in Figure 7 show the three predictions that resulted from these data points and the interpolation between them. In short, "[s]parse sampling in space and time can result in a variety of interpolations from a single data set" (Skole 1994: 442).

Now pair up with a classmate and discuss the following questions:

• So what? Why should we be concerned about sampling, interpolation, and a small number of values? After all, the fewer values, the less calculating I have to do ...
• What would have to be done in order to avoid this problem with few sample points and interpolation for modeling and interpretation? Or: how does the data set have to be improved?

Source: Skole, D. 1995. "Data on land use change: Acquisition, assessment, and analysis." In: Changes in land use/land cover: A global perspective, eds. Meyer, W.B. and B.L. Turner, 444. Reproduced with the permission of Cambridge University Press.

Student Worksheet 2.5

Activity 2.5 Checking for Water-Tightness

• What do you need to ask about the data to assure yourself of their adequate quality (meaning suitable and satisfactory for your intended purpose)?
• What do you need to check?

You may refer back to this list in later exercises and to assess other research articles and the data used in them.
 

___________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________

____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________ ____________________________________________

Student Worksheet 2.6

Activity 2.6 Looking at the Blue Planet with Rose-Colored Glasses

Collect newspaper articles on a recent, much publicized environmental "event," some hot topic that you are interested in, e.g., a devastating earthquake or tropical cyclone, or deforestation in the Amazon or possibly a more local issue. Make a list of all the data provided in each of the articles (e.g., number of deaths, injured people, amount of damage to houses etc., degree of destruction, degree of threat) and discuss when you get back to class why they differ (e.g., because of systematic [political] bias, differences in variable definition, in measurement methodology, scale). To facilitate your overview, you may list these data in form of a table.
 

You may decide to do this activity as a role-playing exercise in which you put on a pair of "rose-colored glasses," i.e., you play someone with a perspective that is slanted in a certain way. For example, you could be the token environmentalist, or corporate executive, or government official. The objective would be for you to try to convince a review panel of scientists (i.e., the rest of the class) of the particular position you take on a chosen issue. Determine your position and develop strategies to convince the panel. (You could do this in a five-minute strategy session with some of the other students.) You will then have a limited amount of time to make your statement to the panel. Emphasize the quality of the data you have on hand to back up your position. After you have all had a chance to make your statements, the panel (class) will vote on which position they found most convincing (most credible, most reliable, most "water-tight").