Geoprocessing Workflow
From objective 2, the geoprocessing workflow on ArcMap required 4 main components to be completed. These components were:
- rate of spread
- host density
- climatic suitability
- cost-distance
Figure 5 shows the general geoprocessing workflow for the entire study, with each component on a different level.

Fig. 5: Geoprocessing workflow to complete the spread analysis of the MPB
Results
1. Rate of spread
Data on the distribution of the MPB was aquired from the Government of Alberta and B.C. Point data was intially collected through aerial surveys and then interpolated into polygon data by forest management authorities.This analysis used data from 2004 - 2017, as 2004 was the earliest year that we had semi-continuous annual data for both Alberta and B.C. These records show rapid expansion of the MPB from the interior of B.C, northeast into central northern Alberta from 2004 - 2010 (Fig. 6). However from 2011-2017, expansion tapered off as the MPB continued to travel farther north and establish more in central Alberta. Fig. 7 shows the total area accumulated of the MPB historical extent from 2004 - 2017. Rate of spread was calculated as ~30 km/yr from 2004-2010, which estimates the MPB could travel a maximum distance of ~2500 km by 2100.

Fig. 6: MPB distribution from 2004-2017 in Alberta and British Columbia

Fig. 7: Total Area (km2) vs. Time (year), derived from MPB distributions in Fig. 6
2. Host density
Data on the percent species compostion and total tree volume was acquired from Beaudoin et al. (2017). They used 2011 MODIS data on the spectral radiance and the k nearest neighbour (kNN) approach to estimate 4 land cover classes, 11 tree stand structure variables, and data on over 70 species of tree at a 250m resolution. Data was combined through addition to create an overall pine tree percent composition layer that included 9 different types of pine (Fig. 8.1):
- Pinus albicaulis (whitebark pine)
- Pinus banksiana (jack pine)
- Pinus contorta (lodgepole pine)
- Pinus monticola (western white pine)
- Pinus ponderosa (ponderosa pine)
- Pinus resinosa (red pine)
- Pinus spp. (unidentified pine)
- Pinus strobus (eastern white pine)
- Pinus sylvestris (scots pine)
Most notably, historic primary hosts of the MPB, lodgepole pine, contributed heavily to the compostion of pine in the interior of B.C., while jack pine is the dominant pine tree within the boreal forest. Tree volume data shows a conistent latitudal trend, decreasing in volume at higher latitudes (Fig. 8.2). Notably, the historic range of the MPB is found in the region with the largest amount of pine tree volume (m3/ha) in Canada (Fig. 8.3).

Fig. 8.1: Pine tree percent composition for Canada

Fig. 8.2: Total tree volume (m3/ha) for Canada

Fig. 8.3: Pine tree volume (m3/ha), product of Fig. 6.1 and 6.2
From Fig. 8.3, we estimated a value for b for Equation 2 of 6.25. Using this value, we derived a resistance equation (Equation 5). Fig. 9 shows the inverse relationship of this resistance function. Fig. 10 shows the final output for the resistance layer, with red areas highlighting areas of high resistance (low pine tree volume) and blue areas of low resistance (high pine tree volume).

Equation 5: Resistance function, b value derived from maximum pine tree volume

Fig. 9: Resistance vs. Pine Volume (m3/ha), b = 6.25

Fig. 10: Resistance of area to MPB infestation, generated using Fig. 8.3 and Equation 5
3. Climate Suitability
Future climate projections were acquired from AdaptWest. Datasets were developed using interpolation methods, specifically Parameter Regression of Independent Slopes Model (PRISM), and data from the Coupled Model Intecomparision Project phase 5 (CMIP5) based on the 5th IPCC Assessment Report. We used an average of 15 low-emission scenario projections for minimum January temperature for 2020, 2050, and 2080. Data was converted from ASCII to raster, then reclassifyed to 5 oC intervals. For this analysis, we are concerned with -40 oC temperatues, as it coresponds with 100% MPB mortality. These projections suggest that by 2020, that only large swaths of Nunavut will reach a minimum of -40 oC (Fig. 11.1). By 2050, only some regions of the Arctic Archipelago, Devon and Ellsemere island, will reach a minimum of -40 oC (Fig. 11.2). By 2080, the minimum temperature is projected to not exceed -39 oC (Fig. 11.3). These projections suggest that:
- the MPB will not be limited by -40oC temperatures in its northern range in the future;
- larger proportions of MPB will survive overwinter and due to warmer temperatures.

Fig. 11.1: Ensemble climate projection of January minimum temperature (oC) for 2020

Fig. 11.2: Ensemble climate projection of January minimum temperature (oC) for 2050

Fig. 11.3: Ensemble climate projection for January minimum temperature (oC) for 2080
4. Cost-distance
To complete the cost-distance analysis, the 2011-2017 distribution data was used as the souce, pine tree volume resistance was used as the cost, and a maximum distance of 2500 km was used. Climate was only not used as a constraint of the final analysis, as it would not have an impact on the results. The final output was also evaluated at a 250 m cell size to fit the same scale as the tree data due to it having the largerst impact on the analaysis. The cost-distance was then reclassified, to fit ~10 year intervals (Fig. 12). This analysis predicts a large expansion, diverging to east as well as the north-west. By 2040, large central regions of Alberta near the Rocky Mountains will have well established MPB populations. Northern B.C., Yukon, and the North-West Territories will also see increasing amounts of MPB. By 2100, the Yukon will see a creep of MPB towards its interior. Alberta and Saskatchewan will see continual radial expansion throughout the boreal forest in its northern regions, with the prairies limiting the MPB range to the south. It can be expected further in the future for the MPB to follow this course, diverging towards Alaska and Northern Ontario.

Fig. 12: Cost-distance surface showing the predicted spread of the MPB until 2100