You have been hired by a local landowner to calculate the carbon sequestration potential of a small forested area in southeastern Michigan. You know it is possible to estimate carbon values using measurements of tree height and diameter. After an initial site visit to the property, you determine it will be too costly and time-consuming to measure every single tree on the property. Given these limitations, you decide to use a representative sample to estimate values for the entire forest. After setting up a sampling plan, you collect information in the field for 18 sample areas. After returning to the office, you enter your data into two CSV files, one with tree measurements and plot identification numbers and another with plot GPS coordinates. You get to use ArcGIS and the Spatial Analyst extension to create a plot shapefile from your tabular data and interpolate your sample data for the entire forest.
At the successful completion of Lesson 6, you will have:
If you have questions now or at any point during this lesson, please feel free to post them to the Lesson 6 Discussion.
This lesson is worth 100 points and is one week in length. Please refer to the Course Calendar for specific time frames and due dates. To finish this lesson, you must complete the activities listed below. You may find it useful to print this page out first so that you can follow along with the directions. Simply click the arrow to navigate through the lesson and complete the activities in the order that they are displayed.
SDG image retrieved from the United Nations [1]
Calculating carbon sequestration and associated carbon credits for forests is an initiative related to climate change. Climate change, also known as global warming, is caused by increased levels of greenhouse gases trapped in Earth’s atmosphere. Some of the expected effects of global warming include melting glaciers, sea-level rise, changes in water resources, changes in food production, loss of biodiversity, increases in extreme weather, and threats to human health.
Forests play an important role in climate change due to the fact that trees naturally absorb and release carbon dioxide during their life cycle. During photosynthesis, they remove carbon dioxide from the air and store it as organic matter in their trunks, branches, foliage, roots, and soils. This process is known as carbon sequestration. When trees decay or burn, they release their stored carbon back into the atmosphere. The amount of carbon a particular tree absorbs or releases throughout its lifecycle is negligible. However, due to their global abundance, the cumulative effect is very large.
The Intergovernmental Panel on Climate Change (IPCC) [2], winner of the 2007 Nobel Peace Prize, is the United Nations body for assessing the science related to climate change and, therefore, often considered the world’s leading authority on climate change. One of their recommendations is that we need to mitigate the future impacts of climate change by reducing current and future emissions. One way to reduce emissions is to reduce the number of forests that are clear-cut or degraded since these activities account for about 15% of global greenhouse gas emissions. There are simply some existing natural environments that we cannot afford to lose due to their irrecoverable carbon reserves. And, preserving existing forests is considered a much better alternative to reforestation or afforestation, as it takes decades for a new tree to grow and absorb the amount of carbon that is released when a mature tree is lost.
There are several initiatives to encourage the preservation of existing forests on a global scale. Reduced Emissions from Deforestation and Forest Degradation (REDD+) [3] is a framework by the UNFCCC (United Nations Framework Convention on Climate Change) Conference [4] that guides activities that will provide financial compensation to landowners for maintaining and protecting forests. To receive compensation under the REDD+ initiative, landowners need to be able to assess the amount of carbon that is stored in the trees on their property. Many of the methods to do this require a forest inventory in which tree species, height, and diameter at breast height (DBH) are measured. This information is used to estimate the volume of organic matter for each tree, which is then translated into results-based financing, whose value fluctuates depending on the current market trading value of carbon.
While it would be more accurate to measure every tree in a forest during field inventories, limitations of time and money typically make this unfeasible. This is especially true for large or hard-to-access areas (e.g., mangrove forests, swamp forests, forests with steep topography). Therefore, a common practice has been to collect a representative sample and then interpolate the values for the entire study area or use remotely sensed data. In this activity, we will concentrate our efforts on a representative sample. Forests can be inventoried by demarcating a number of sampling locations known as plots. Trees are only measured if they fall within the plot boundaries. The number and location of plots required for a given area depend on the size of the forest and the amount of variation within the study area. Large forests or forests with a lot of variation in tree cover will require more plots than small forests or forests with uniform tree cover.
GIS can be very helpful when trying to decide on the number and location of sample plots. For example, you can overlay your site boundary with current and historical aerial photos to look for variations in forest cover. You can also incorporate other datasets such as Digital Elevation Models (DEMs), hydrology, parcel boundaries, and roads. These datasets will help you identify possible hazards in the field such as fences, large rivers, steep terrain, etc. They can also help you estimate the age of the forest, depending on how old the forest is and how far back you can find aerial photos.
Most of these concepts are not unique to forest inventories. It is actually quite common in the environmental field to use representative samples to understand larger areas. For example, environmental consultants typically install monitoring wells to understand how groundwater and soil conditions vary across a site. By measuring water levels in the monitoring wells, they are able to calculate the direction and speed of groundwater flow for the whole site. They also collect groundwater and soil samples at these locations to determine if pollution levels exceed legal limits. Since the locations of the samples are known, it is possible to plot them on a map and use their location to predict values between monitoring wells.
After data is collected in the field, it is common to enter the data into an electronic format such as an Excel table, CSV file, or a simple database. As long as the field measurements are in a digital format, it is possible to view them in ArcGIS. Depending on the native file type, you may need to complete some intermediate processing steps for ArcGIS Pro to recognize them. Using the "Join" and "Display XY Data" tools in ArcGIS, it is possible to create shapefiles from tabular data. You can then use your shapefiles to identify spatial patterns in your data and create maps of your results.
The interpolation tools available within the ArcGIS Spatial Analyst Extension are particularly helpful when working with representative point data. In Lesson 6, we will practice interpolating point data to estimate values for areas that were not actually sampled. We will also explore how the options and environment settings listed below affect the output grids created by Spatial Analyst tools. We will read more about these settings in the help articles listed in the required readings section.
The required readings for Lesson 6 are listed below. Each of the short articles provides specific information about the tools and techniques we will use in Lesson 6.
Find the help articles listed below on ArcGIS Pro Resources Center [5] website.
This section provides links to download the Lesson 6 data along with reference information about each dataset (metadata). Briefly review the information below so you have a general idea of the data we will use in this lesson.
Note: You should not complete this activity until you have read through all of the pages in Lesson 6. See the Lesson 6 Checklist for further information.
Create a new folder in your GEOG487 folder called "L6." Download a zip file of the Lesson 6 Data [13] and save it in your "L6" folder. Extract the zip file and view the contents. Information about all datasets used in the lesson is provided below:
The data we will use In Lesson 6 was collected by the International Forestry Resources and Institutions (IFRI) Organization [14]. IFRI is a research network made up of 18 collaborating research centers around the globe. Since 1992, IFRI researchers have collected both ecological and social field data for over 400 sites in 15 countries. For this lesson, we are going to use a subset of their data to calculate the carbon sequestration and carbon credits for a small forest located in southeastern Michigan.
The data was collected by laying out 18 circular plots, every 10 meters in diameter, at random locations throughout the study forest. The coordinates of these locations were determined in advance using GIS. In the field, students used GPS and mobile apps to navigate to the middle of each plot, lay out the circular plot, and collect attribute information about the trees. Any trees that fell inside the plot boundaries were measured to obtain their height and diameter. Measurements were collected for a total of 278 trees.
The Step-by-Step Activity for Lesson 6 is divided into two parts. In Part I, we will create a point shapefile from our starting data tables. We will then use the field calculator to calculate the carbon sequestration for each tree and create totals by plot. In Part II, we will use the Spatial Analyst extension tools to interpolate the plot data to a raster grid covering the entire study area. We will use the results to calculate the total carbon for the study forest. During interpolation, we will experiment with several different toolbar settings to see how they affect the results.
Note: You should not complete this step until you have read through all of the pages in Lesson 6. See the Lesson 6 Checklist for further information.
In Part I, we will create the shapefile we will use to interpolate our data (a point shapefile of plots with the total carbon as an attribute). To create this, we start with the two CSV files "GPS.csv" and "Tree _Measurements.csv".
How far is the study forest from the city of Ann Arbor, MI or State College, PA? What is the surrounding land used for (commercial, agriculture, residential, etc.)?
Make sure you have the correct answer before moving on to the next step.
Check the Properties > Source Tab > Spatial Reference to make sure the Plot shapefile was projected correctly to NAD 183 UTM Zone 16N. If your projection doesn’t match, make sure you remove the base maps, and choose the coordinate system of the Map.
Make sure you have the correct answer before moving on to the next step.
Check the location of your plots by comparing your plot shapefile to the map below. Note: Your map will not look exactly like this by default. I changed the symbology of the points, added labels of Plot ID's, and added the Imagery layer in the background to make it easier to compare your data to the example. If you add the imagery base map again, make sure you remove it from your map and Save before moving on to the next step.
We are going to use a somewhat general set of equations to estimate the carbon stored in each tree. For this lesson, we do not need a high level of accuracy. The important part is to demonstrate the concept of how one can calculate carbon credits using GIS. You can read more about the method we will use at: How to calculate the amount of CO2 sequestered in a tree per year [16].
There are more sophisticated methods you can use that take into account the tree species, age, climate, and other factors. The paper, “Methods for Calculating Forest Ecosystem and Harvested Carbon with Standard Estimates for Forest Types of the United States [17]” highlights an example of a more complex methodology. An example of a simpler method is highlighted in the “Landowner’s Guide to Determining Weight and Value of Standing Pine Trees [18]”.
Variable | Description | Units | Equation |
---|---|---|---|
D | Measured tree diameter (DBH) | Inches | See Tree Measurements Table (be careful with your units here). |
H | Measured tree height | Feet | See Tree Measurements Table (be careful with your units here). |
Wa | Total above-ground weight of the tree (w/o roots) | Pounds | Wa = 0.15D2 *H |
Wt | Total weight of the tree and roots | Pounds | Wt = 1.2 Wa |
Wd | Dry weight of the tree | Pounds | Wd = 0.725Wt |
Wc | Weight of carbon in the tree | Pounds | Wc = 0.5Wd |
Ws | Weight of carbon dioxide sequestered in the tree | Pounds | Ws=3.6663Wc |
Make sure you have the correct answer before moving on to the next step.
Compare your data with the summary statistics below for the "Ws" variable.
Mean | 801.0505393089 |
---|---|
Median | 337.511190219 |
Std. Dev. | 1,171.7755087661 |
Count | 278 |
Min | 0 |
Max | 6,748.03406916 |
Sum | 222,692.04992788 |
Nulls | 0 |
Skewness | 2.6994064237 |
Kurtosis | 10.3340354971 |
If your data does not match this, go back and redo your calculations. Pay special attention to unit conversations (make sure to round to the nearest 4 decimal places), data types of the fields you used, and typos in equations.
Make sure you have the correct answer before moving on to the next step.
Compare your data with the summary statistics below for the "c_lbsqm" variable.
Mean | 157.6023000198 |
---|---|
Median | 109.55277128 |
Std. Dev. | 162.8283983546 |
Count | 18 |
Min | 6.420046445 |
Max | 646.682030534 |
Sum | 2,836.8414003566 |
Nulls | 0 |
Skewness | 1.5234771786 |
Kurtosis | 5.4280093282 |
If your data does not match this, go back and redo your calculations.
In Part II, we will use the Spatial Analyst extension tools to interpolate the carbon sequestration data we calculated for each plot to the entire forest. We will run the same interpolation tool several times to see how altering the extent, mask, and cell size settings affect the results. We will start by accepting all default settings. Then we will change the settings one at a time to see how each one affects the results.
Do some plots have more trees than others? Is there a lot of variation in the total amount of carbon or carbon per square meter value? If so, why do you think this may occur? Hint: Look at an aerial image basemap.
Do you see any spatial patterns in the data? For example, do some areas of the forest have higher values than others? If so, why do you think this may occur?
Remember from the Background Information section that the Spatial Analyst tools are governed by user-specified settings. Two of the most common errors when using Spatial Analyst tools are to either completely ignore these settings, or to set them improperly. Let’s try to interpolate our data using all of the defaults and see what our results look like.
Make sure you double-check ALL environment settings before running ANY tools in Spatial Analyst! The program often resets your cell size, extent, and mask to program or data layer defaults.
Click the Show Help >> button to help define particular input parameters.
You can review the specific input and environment settings you used in the Analysis tab, Geoprocessing group > History. This can be helpful if you are not sure if you made a mistake somewhere along the way during a complex workflow.
Make sure you have the correct answer before moving on to the next step.
If your map does not match the example below, go back and redo the previous step.
What is the default setting for analysis extent?
What is the cell size of the "default" raster we created? Why?
Raster Attribute Tables
You may notice that the option to open the attribute table of the "default" raster is grayed out. ArcGIS Pro only builds raster attribute tables if certain conditions are met. One of the conditions is that the values in the raster have to be integers. Since the values in our raster have decimals, it is not possible to view the attribute table.
Now that we’ve explored the default settings, let’s see what happens if we alter just the extent settings. Unlike vector files, rasters will always have the shape of a perfect rectangle. The size and location of the rectangle is defined by its extent.
Make sure you have the correct answer before moving on to the next step.
If your map does not match the example below, go back and redo the previous step.
How do the extents of the "parcel_extent," "forest_extent" and "default" rasters compare?
Now, let’s see what happens if we alter the mask and extent settings. Even though all rasters are defined as perfect rectangles, you can still represent your data as a sinuous shape. The computer creates this illusion by assigning cells outside the sinuous shape values of "NoData." There is not a direct equivalent to this concept in vector files.
What would the output raster look like in the following scenario?
In Step 3, we learned that the default cell size will depend on the input data. If you are using one or more rasters as inputs, the cell size will default to the coarsest raster resolution. If you are using a vector file, it will calculate the cell size based on the extent of the file to create 250 cells. The default for rasters seems appropriate since GIS best practices dictate that you should always go with the cell size of your coarsest input dataset. However, the default for vector files is quite arbitrary.
How do we choose a more meaningful cell size for our analysis? One rule of thumb is that you don’t want to "create" higher resolution data than what exists in your measured values. We know that the tree data was collected by measuring trees that fell within 10 m diameter circular plots. A cell size of 1 cm would not be appropriate, because we do not know how the data varies at that scale. A cell size of 1,000 m would be too large, since it is larger than our study area. For this project, we will use a cell size of 1 m, since our carbon values are in pounds per square meter.
Now that we have an understanding of how the spatial analyst environment settings function, we can return to our original question. We want to figure out how much carbon the study forest sequesters. To accomplish this, we will use the "Zonal Statistics" tool in Spatial Analyst. This tool allows us to calculate statistics of the cell values of one raster (e.g., carbon1m) within zones specified by another file (e.g., forest boundary). We will use it to sum the carbon values in each cell to create a total for the entire forest.
ArcGIS may not show all the digits in a table by default. If your numbers do not match the numbers in the quiz, expand the columns in your table to display all the digits.
The monetary value of each carbon credit fluctuates based on the current market conditions. Check out the more information about the California’s cap and trade system at Center for Climate and Energy Solutions(C2ES) [20].
One of the main take away points from this lesson is that Spatial Analyst is a modeling tool. Models don’t give exact final answers; rather they give you estimates of reasonable answers based on a set of assumptions.
Environments settings allow you to easily alter the underlying assumptions of your model (cell size, mask, extent) and then quickly recalculate your results.
Selecting environment settings in Spatial Analyst tools can be confusing and seem somewhat arbitrary. If you don’t know which Environments setting you should use for a particular scenario, you can try experimenting with a variety of options. This type of sensitivity analysis will help you understand how changing model assumptions affect your final results.
That’s it for the required portion of the Lesson 6 Step-by-Step Activity. Please consult the Lesson Checklist for instructions on what to do next.
Try one or more of the activities listed below:
Note: Try This! Activities are voluntary and are not graded, though I encourage you to complete the activity and share comments about your experience on the lesson discussion board.
Advanced Activities are designed to make you apply the skills you learned in this lesson to solve an environmental problem and/or explore additional resources related to lesson topics.
After looking at current aerial photos of the site, you realize that the density of tree cover varies across the study area. In fact, there are many parts of the study area that are not covered with trees. Historical aerial photos reveal that the forest is quite young and that many parts of the study area were used for agriculture until quite recently. Using aerial photos from the 1940s to the present, you create a shapefile showing areas of similar age and forest type (Vegetation07.shp in your L6 folder).
Given this new information, you realize that your original estimate of carbon sequestered by the forest is too large since many of the interpolated cells are located within areas that are not covered with trees. You decide to redo your interpolation to remove areas that are not forested from your results. What is your new estimate of carbon sequestered by the study forest?
In Lesson 6, we talked about climate change, forests, and carbon credits. We explored how to use GIS to plot and interpolate representative sample data measured in the field. We also explored how altering the extent, mask, and cell size settings within Spatial Analyst lead to very different results. We learned that you need to select settings appropriate for your analysis since accepting the defaults can have unintended consequences. In the next lesson, we will look at two different Spatial Analyst tools: "reclassify" and "tabulate area" to investigate land use change over time.
Lesson 6 is worth a total of 100 points.
If you have anything you'd like to comment on, or add to the lesson materials, feel free to post your thoughts in the Lesson 6 Discussion. For example, what did you have the most trouble with in this lesson? Was there anything useful here that you'd like to try in your own workplace?
This page includes links to resources such as additional readings, websites, and videos related to the lesson concepts. Feel free to explore these on your own. If you would like to suggest other resources for this list please send the instructor an email.
Links
[1] https://www.un.org/sustainabledevelopment/news/communications-material/#FAQ
[2] https://www.ipcc.ch/
[3] https://unfccc.int/topics/land-use/workstreams/redd/what-is-redd#
[4] https://unfccc.int/
[5] https://www.esri.com/en-us/arcgis/products/arcgis-pro/resources
[6] https://pro.arcgis.com/en/pro-app/help/analysis/spatial-analyst/basics/what-is-the-spatial-analyst-extension.htm
[7] https://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/an-overview-of-the-spatial-analyst-toolbox.htm
[8] https://pro.arcgis.com/en/pro-app/help/analysis/spatial-analyst/performing-analysis/the-analysis-environment-of-spatial-analyst.htm
[9] https://pro.arcgis.com/en/pro-app/help/analysis/geoprocessing/basics/geoprocessing-environment-settings.htm
[10] https://pro.arcgis.com/en/pro-app/tool-reference/geostatistical-analyst/an-overview-of-the-interpolation-toolset.htm
[11] https://pro.arcgis.com/en/pro-app/help/analysis/geostatistical-analyst/how-inverse-distance-weighted-interpolation-works.htm
[12] https://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/an-overview-of-the-zonal-tools.htm
[13] https://www.e-education.psu.edu/geog487/sites/www.e-education.psu.edu.geog487/files/image/lesson06/Pro/L6_Data.zip
[14] http://ifri.forgov.org/
[15] https://www.e-education.psu.edu/geog487/sites/www.e-education.psu.edu.geog487/files/image/lesson05/Pro/Lesson5ExportTable.png
[16] https://www.e-education.psu.edu/geog487/sites/www.e-education.psu.edu.geog487/files/activities/lesson05/Calculating_CO2_Sequestration_by_Trees.pdf
[17] https://www.fs.usda.gov/ecosystemservices/pdf/estimates-forest-types.pdf
[18] https://www.uaex.uada.edu/publications/pdf/FSA-5017.pdf
[19] http://www.onlineconversion.com
[20] https://www.c2es.org/content/california-cap-and-trade/
[21] https://www.fao.org/redd/areas-of-work/national-forest-monitoring-system/en/
[22] http://unfccc.int/2860.php
[23] http://www.ipcc.ch/
[24] https://www.ipcc.ch/reports/
[25] https://www.fs.usda.gov/ccrc/
[26] https://unfccc.int/climate-action/united-nations-carbon-offset-platform?gclid=CjwKCAjwh4ObBhAzEiwAHzZYU2suPgbJ_x7pI0m4SS7O4jBSQAd1uhcH3Rdj3kf8wB-coLkOqsoxwRoCj_gQAvD_BwE
[27] https://www.greenclimate.fund/redd
[28] http://www.treebenefits.com/calculator/
[29] https://www.psu.edu/news/research/story/research-help-private-forest-owners-manage-woodlands-ecosystem-services/
[30] https://rdcu.be/dmDiZ
[31] https://www.ipcc.ch/working-group/wg2/?idp=158
[32] https://www.ucsusa.org/sites/default/files/2019-10/deforestation-success-stories-2014.pdf
[33] http://www.uncclearn.org/wp-content/uploads/library/13052015unepviet2.pdf
[34] https://crsreports.congress.gov/product/pdf/R/R46952