Forests cover about 30% of the earth's land surface and are an important resource for humans and ecosystems, as they provide a variety of ecological, economical, and social services (Bonan, 2008). Forests also play an important role in mitigating climate change, as they absorb carbon from the atmosphere (Bonan, 2008). Carbon dioxide is a greenhouse gas emission that has increased in concentration in the atmosphere since the pre-industrial era, leading to changes in the earth's weather and climate, including global warming, sea level rise, and ocean acidification (IPCC, 2014).
Figure 1 shows the percent of land covered by forest in various geographic regions of the world between 1990 and 2010. Many regions appear to be either fairly stable or experiencing deforestation, which reduces carbon storage.
Data Source: African Development Bank Group
Figure 1
Figure 2 shows carbon emissions, measured in kilotons, by geographic region. Carbon emissions are steadily increasing for each geographic region. This increase in carbon emissions, in combination with the decrease of forest land shown in Figure 1, means an acceleration of climate change if no mitigation strategies are implemented.
Data Source: African Development Bank Group
Figure 2
The purpose of this project was to derive metrics from LiDAR data that could be used for forest carbon inventories. A study by Stephens et al. (2012) explored the potential for using different LiDAR derived metrics to analyze forest carbon inventory, rather than costly ground inventories. The authors found that the two best variables for predicting carbon stock were LiDAR 30th percentile height and percent crown volume. The goal of the present analysis was to use derive forest canopy cover and tree height metrics (used to determine 30th percentile height) from LiDAR data for use in a future forest carbon inventory.
Figure 3
The data for this analysis covers a section of the Capitol State Forest, which is southwest of Olympia, Washington (see Figure 3). This forest is managed by the Washington State Department of Natural Resources and is used primarily as a resource for timber and recreation, but also naturally serves as carbon storage (Ear to the Ground, 2019). The LiDAR data was downloaded from the Washington State Department of Natural Resources Washington LiDAR Portal. This section of the forest was surveyed in 2012, so the data is eight years old. Because timber harvest occurs frequently in this forest, data from 2012 likely will not reflect the forest’s current metrics. However, resolution was 10.94 points per square meter, which exceeded the target of 8 points per square meter and was filtered to remove inaccurate points (Martinez, 2012).
The analysis for this project resulted in two rasters: canopy cover and tree height. To create the canopy cover raster, I followed ESRI’s Estimating forest canopy density and height tutorial. To use the LiDAR data, the point cloud LAZ needed to be converted to LAS format. This was accomplished using an open source software called LASzip. Next, the data was separated into ground (class 2) and vegetation (classes 1, 3, 4, and 5) layers, including all return types, using the Make LAS Dataset Layer tool in ArcGIS Pro. The layers were converted to rasters and null values converted to equal zero. Finally, the ground raster was divided by the sum of the ground and vegetation rasters to find the vegetation density above ground (see Figure 4). In order to find the total canopy cover percentage for the area of interest, this raster was converted to integer format, an attribute table was built for the layer, and field calculator was used to calculate the percent of cells containing vegetation.
Figure 4
To create the tree height raster, I followed Earth Lab’s Canopy Height Models, Digital Surface Models & Digital Elevation Models – Work With LiDAR Data in Python tutorial. To complete this step, I downloaded the DSM and DTM data from the Washington LiDAR Portal for my area of interest. The measurement unit for this data was in feet. To create the raster, I simply used the Minus tool in ArcGIS Pro to subtract the DTM layer from the DSM layer. The output was a raster (see Figure 5) ranging in height from zero to 273 feet.
Figure 5