The Value of Landanderson & Sutton

The Value of LandAnderson & Sutton

The Value of Land

Global & National Valuation of Ecosystem Services

― Sharolyn J. Anderson and Paul C. Sutton

Introduction

Problem

In 1997 a seminal paper in Ecological Economics was published in the journal Nature. The title of the paper was: The value of the world’s ecosystem services and natural capital. The key findings were an estimate of the global value of ecosystem services measured in dollars. The values for ecosystem services were derived from hundreds of papers from which estimates of the dollar value of ecosystem services were derived for various ecosystems associated with various land cover classes. The total global value at that time was $33 Trillion dollars per year. This figure was roughly double the size of the entire marketed global economy. Ecosystem services are analogous to the interest on natural capital. Thus, this estimate of $33 Trillion dollars is not the total value of the earth’s land cover but the value of the interest provided by the services that are derived from this land cover annually. Consequently it is very interesting to compare the relative magnitude of the annual value of ecosystem services to the size of the global market economy in a common unit – the dollar.

It is now 2015 and an update to these estimates of ecosystem service values are warranted. The estimates will change for several reasons including: 1) improved estimates of ecosystem value, 2) changes to areal extent of various global land covers, 3) the size of the estimate relative to the market economy may also change.

1)What data set or sets best represent the marine and terrestrial cover of the entire earth?

2)What is the state of the art dataset for representing ecosystem service values?

3)What is the current estimate of the total value of the earth’s ecosystem services?

4)How can we subset this data to answer questions like this at a national scale?

Location

The geographic scope of this laboratory exercise is global and national.

Time to complete the lab

Completion of this laboratory exercise should take 1-3hours.

Prerequisites

An understanding of elementary GIS (preferably ArcMap), and basic statistics and mathematics.

Keywords

Topical: (Land cover, Land use, mixed pixel, MAUP, Ecosystem Services, Ecosystem service valuation, benefits transfer model).

GISc: Concepts / tools / techniques used (raster calculator, cell based modeling, Table Joins, Field Calculations, Table Sorting) and types of data used (raster and vector).

Data used in this lab

National boundaries of the world’s countries (ESRI)

Global Ecosystem Data (Modified GlobCOV )

Spatial Reference:World Mollweide

Datum: D_WGS_1984

Units: meters

Data must have no copyright restrictions; otherwise, please provide data sources and permissions using the ESRI Press Copyright Permissions Form.

Student activity

One difficulty of studying global land cover change is consistency of datasets. Despite the fact that Landsat, MODIS, AVHRR, and other satellite systems have been functioning for decades there are few consistent and comparable global land cover composite datasets from which measures of land cover change can be made. (answer Q1).

The Economics of Ecosystems and Biodiversity (TEEB) is a global initiative on “making nature’s values visible” ( ). The TEEB project in collaboration with the Ecosystem Services Partnership (ESP ) houses an Ecosystem Service Valuation Database that is available on the web to all interested parties. This dataset summarized over 300 case studies that results in over 1,350 data points with monetary values of ecosystem services. At the global scale an exercise in aggregation and consolidation is necessary in order to perform a global assessment of ecosystem value. We will use a ‘benefits transfer’ model for ecosystem service valuation. The Earth Economics Institute provides a clear definition of the benefits transfer method here:

“Benefit transfer involves obtaining an estimate for the value of ecosystem services through the analysis of a group of primary valuation studies which have been previously carried out to value similar goods or services in similar geographies and contexts. The transfer itself refers to the application of derived values from the original study site to a policy site. As the "bedrock of practical policy analysis", benefit transfer has gained popularity in the last several decades as decision-makers have sought timely and cost-effective ways to value the ecosystem services of natural capital”. ( )

Our dataset is a composite of several datasets because no single dataset was designed to map ecosystem service on a biome basis. We needed a global dataset with the following terrestrial and marine classificationsand their associated Ecosystem Service Values (ESV).We found that the best baseline dataset for our purposes was GlobCOVER (Figure 1). However, the land cover categories for GlobCOVER are only terrestrial and are vegetation based rather than biome based. We aggregated the GlobCOVER land categories according to biomes of the TEEB table of ESVs (Table 1).

Table 1: Ecosystem Service Values by Biome distilled from TEEB ESV database

We were particularly unsatisfied with the category “Artificial surfaces and associated areas” so we used a state-of-the-art global human settlements layer to define the urban or built areas of the globe ( ). We also created a ‘Tundra’ class by converting grasslands between 60° and 75° north. Coral reefs were added using a ReefBase dataset ( ) as were shelf and marine defined from a global bathymetry dataset.

Figure 1: GlobCOVER data product 2009( )

Table 2: Aggregation of GlobCOVER categories to the TEEB ESV table categories

Exploring the Ecosystem Service Value Data

Start ArcMap by either clicking on the icon on the desktop or by going to ‘Start’ > ‘All Programs’ > ‘ArcGIS’ > ‘ArcMap’. Create a new map document on opening. You can save your work so far as a ‘Map Document’ by choosing file -> Save As -> … browse to a directory and name your file something simple like ‘ESV_exercise.mxd’. Now add the following layers using the ‘add data’ button . in the toolbar. []. Add the following three datasets from the geodatabase named ValueOfLand.gdb : ESV_Base (a raster dataset), World_Nations_2006 (a polygon vector file), Biome_ESV_LCcode (a table) and ESV_Base.lyr (a legend layer). If you are successful in reading inthese data you should be looking at something like the image below (Figure 2).

Figure 2: ArcMAP interface with ESV_Base data displayed.

Again, you can (should ) save your work so far as a ‘Map Document’ by choosing file -> Save As -> … browse to a directory and name your file something simple like ‘ESV_exercise.mxd’. This map document contains one raster layer of the earth’s marine and terrestrial biomes and one vector layer of the world’s national boundaries, and a table that will be joined to link biomes with ESValues. Now is a great time to explore your data. Pan and zoom around the world to study the patterns of landcover.Look for places where you know well and do a ‘1 minute error check’ on the data (answer Q2).

Right click on the layer titled: ESV_Base and choose ‘Open Attribute Table’. Thisopens the ‘Value Attribute Table’ or VAT of the ESV_Base raster layer. The ‘Value’ column is the pixel or grid cell value whereas the ‘count’ column is the number of cells with that particular ‘value’. So, upon inspection of the table you should be able to see that there are 14 different values representing the 14 different biomes. Remember, because we are in a Molleweide projection (i.e. equal area) all the pixels are roughly 1 km2 in areal extent. Therefore we can count pixels to measure areal extent. The category ‘Urban’ has value ‘4’. The category ‘marine’ has value ’13 & 14’ (shelf and ocean). Use this information and the other information in the table to answer question 3. (Answer Q3)

Estimating the Global Value of Ecosystem Services in Dollars

Step 1 - Table Join and Field Calculation:The first approach is to simply do all of the problem within the tables you have already added to the table of contents (TOC). Let’s explore the table Biome_ESV_LCcode. Right Click on Biome_ESV_LCcode and choose ‘Open’. You should get a table to open like the one below. Note: You might have to broaden the column labled ‘ESV’ to see full depth of the numbers. Check to see that your ‘Coral Reefs’ have an ESV of 352249. There are four fields in this table. ‘OBJECTID’ is internal record number that you should NEVER manipulate. In this case it matches the LC_code value but that is not always the case. The other three fields are: Biome, ESV, and LC_code. This is basically a lookup table for the pixels of the ESV_Base raster dataset.

Remember the ESV_Base VAT only has VALUE and COUNT fields. By joining these two tables we can calculate the global value of ecosystem services using a Benefits Transfer approach. Here we go.

Right Click on ESV_Base in the table of contents and choose ‘Joins and Relates’ -> ‘Join’.

This should open a dialog box like the one portrayed below (Figure 4). When you click ‘OK’ you will have joined the second table to your first table. This new ‘joined’ table will now have seven columns titled

‘VAT_ESV_Base.OBJECTID*’, ‘VAT_ESV_Base.Value’,’VAT_ESV_Base.Count’,’VAT_ESV_Base.LCcode.OBJECTID*’,‘Biome’,’ESV’, and ‘LC_Code’. Use the ‘Options’ > ‘Export’ choice to write a table to your workspace. You can ‘edit’ this exported table easily (you can name it whatever you like but ArcMap defaults to names like Export_Output_#.dbf).

At this point it is worth a think about what the meaning of the fields in the table are. Let’s consider the first record or row of the table 4. Four columns have identical values of 1-14 which represent the internal record numbers and the pixel value for each Land Cover Code. The first row of the third field ‘VAT_ESV_Base.Count’ is the number of pixels that have an LC_Code value of 1. Because these pixels are all 1 km2 it means there are 125,175 square kilometers of Coral Reefs on the whole planet. The 5th column provides a meaningful name for the numeric LC_Code code value – ‘Coral Reefs’ in this case. The 6th column ‘ESV’ represents the dollar value of the corresponding biome in dollars per hectare per year.

Export your version of Table 4 (below) to a dbf file that you can read into Microsoft Excel. How can you use this table to calculate the total global value of ecosystem services with some simple spreadsheet tricks?

(Answer Question 4)

Using ArcGIS to calculate the total global value of ecosystem services: Using your version of Table 4 (above) create a field (‘Add Field . . . from the upper left pull down menu). Name the field ‘Biome_ESV’ and give it a Type of ‘Double’ (Using Add field dialog box to the right). Right click on the field name ESV_Biome (VAT_ESV_Base.ESV_Biome) and choose ‘Field Calculator’. Recall that 1 km2 is equivalent to 100 hectares when you answer the next question. (Answer Question #5)

As you perform the field calculation you will likely get a warning about performing a calculation ‘outside of an edit session’. Click ‘yes’ this is not an issue. The ‘field calculation’ dialog box should look like the figure on the right when you do this correctly.

Once you click ‘OK’ you should get a field in the table with the name: VAT_ESV_Base.ESV_Biome that has large value in it. For example, Grasslands (10) should have a value of: 18390437475800.

This means that the total value of ecosystem services derived from grasslands annually for the entire planet is:

$ 18,390,437,475,800

Or roughly 18.4 Trillion dollars per year. (Note: The value for marine biomes in this dataset are not particularly accurate because of scaling and resampling issues.) For a global total of the value of all the world’s ecosystem services you will need to sum all number in the ESV_Biome column. This is done in ArcGIS quite easily by right clicking on the field name and choosing ‘statistics’ (Figure 6).

(Answer Question #6 & #7)

Mapping and Monetizing ESV at the National Level

Now we will choose a particular country and map the biomes and associated terrestrial ecosystem services and values for that country. Please use a country of your choice. We will demonstrate the steps using Zimbabwe.

Select your country (Zimbabwe) from the World_Nations_2006 layer. There are many ways to do this. Since this is a single country selection and we know where it is we will choose Zimbabwe by using the select features tool . and simply ‘click’ on Zimbabwe (Figure 7).

Figure 7: ArcGIS window with ‘Zimbabwe’ selected as indicated by CYAN border.

Now right click on the World_Nations_2006 layer and choose ‘Selection -> Zoom to selected features’. Make sure this is the country you want to map.

We want to extract the pixels for the selected country. Remember we want to have a VAT table that represents the counts of pixels for that country only.

We perform the ‘clipping’ operation or ‘extraction’ using the ‘Extract by Mask’ tool in the Arc Toolbox Spatial Analyst Tools-> Extraction -> Extract by Mask (Figure 8)

NOTE: Make sure that your country is still ‘selected’ (e.g. highlighted in the Cyan Blue color).

Figure 8: Dialog box for ‘extract by mask’

After clicking buttons like ‘yes’, ‘close’, and ‘ok’ you should end up with a raster layer added to your TOC that is named after the country you selected. Uncheck the ESV_Base in the TOC zoom to your country. We need to make sure that the VAT table for your specific country has been recalculated. We do this using the ‘Build Raster Attribute Table’ Tool. Please click on the search button and type: ‘Build Attribute’. Select the second item on the found list.

Figure 9: Build Raster Attribute Table Dialog box

Do a one minute error check on your country’s representation. (Answer Question #8)

After you verify that your country is good to go, proceed with the same operations that you did for the entire world on your single country dataset (Page 6).

Figure 10: ArcGIS window with National Level Analysis for Zimbabwe

1)Join the Biome_ESV_LCcode table to your Country VAT

2)Add and caculate the ESV_Biome field in the joined table.

3)Summarize the ESV_Biome field for your country using the ‘statistics’ function.

PRoducing a Powerpoint Slide summarizing your GIS work

To produce a powerpont slide for your country you will likely want to have various images and tables representing the important and salient information. There are various programs such as ‘Snag-IT’ and other screen capturing utilities that allow you to grab images right out of ArcGIS. In addition ArcGIS allows you to export tables, JPGs, and PNGs, in addition to maps. Arc has a layout functionality that many people like to use. On the following page we present a single slide powerpoint for the nation of Zimbabwe that was made using a few screen captures, manipulating tables in excel, and the draw functionality of Powerpoint.

Create a similar powerpoint slide for your country. We find this to be the fun part of the job (Answer Question#9)

Figure 11: A sample powerpoint slide summarizing Ecosystem Service Value for Zimbabwe

Answer and/or Solve the Following Questions and Problems

Q1: Do some internet searching for global land cover datasets. Find a minimum of three and provide an image and URL for each of them. What problems would you have using these to measure land cover change? Hints: Classification scheme, availability in time series, etc.

Q2: Does the data have a ‘ring of truth’ to it? In other words do the areas of the world that you know appear as you would expect them to in this dataset? Provide an example of your ‘one minute error check’.

Q3: Calculate the percentage of the terrestrial earth’s surface that has been urbanized (e.g. covered with built up area by humans) according to this ESV_Base dataset. Does this ring true to you? Do a quick internet search to cross-check the validity of your finding?What is the general consensus as to the fraction of the earth’s land that has been urbanized? Does this number surprise you? Why?

Q4: How could you use a spreadsheet program like Microsoft Excel to take this table and calculate the total value of the world’s ecosystem services? (Note: We will use ArcGIS to do the same thing)