1. Utilizing Landsat TM and Forest Service Aerial Survey Data for Mapping Mountain Pine Beetle Outbreak in Medicine Bow National Forest, WY UW Undergraduate Research Day Presentation by Paul A. Arendt w/ Dr. Ramesh Sivanpillai, 24 April 2009 Photo from July 2008, by author

2. - Mountain pine beetle outbreak has been expanding for the past 12-14 years in the Rocky Mountain region - Medicine Bow National Forest: 256, 100 affected acres in 2007 460,300 acres by 2008! [1] - Carbon County: 238,000 acres in 2007 365,000 acres in 2008 [1] Image courtesy of http://www.fs.fed.us

3. Image courtesy of http://slf-web.state.wy.us

4. Image courtesy of http://www.fs.fed.us

5. Management issues: Falling trees near power lines, roads, etc. Impacts on wildlife and watersheds Forest fire management Forest Service’s response: Bark Beetle Incident Implementation Plan 2007 – 2011: Vegetation treatments (i.e., logging, fuels reduction projects) Costly to monitor the growth of the outbreak using only field surveys however…. - Image courtesy of http://www.fs.fed.us

6. - Remote sensing: Studying the Earth from a distance (aerial photography and satellite imagery) Materials reflect and absorb EMR in different ways (vegetation has a strong red/IR pattern) - Remotely sensed data can help land managers make more informed decisions with regards to mitigating hazardous conditions and better allocating resources - Additionally… Quicker than field work on the ground Opportunity to explore outbreak’s environmental drivers Image courtesy of http://www.crisp.nus.edu.sg/

7. - In the Medicine Bow National Forest, are beetle impacted sites exhibiting a different reflectance pattern than non-impacted areas? NDVI – photosynthetic activity Wetness – 3 rd Tasseled Cap Transformation, measures amount of moisture: Negative values are dry Positive values are wet - If so… Does the rate of decline vary according to different elevation zones? Does damage vary according various slope aspects?

8. Image courtesy of http://www.fs.fed.us

9. - Landsat TM 5: Archive made available in January 2009 Moderate spatial and spectral resolution: - 30 m pixel size (~ 100 ft) - Shuttle Radar Topography Mission (SRTM) Flown in 2000 SAR (Synthetic Aperture Radar) Accurate w/ 1 arc second (30 m) res. Image courtesy of wikipedia.org Image courtesy of http://www.fas.org

10. Shuttle Radar Topography Mission (SRTM ) dataset

11. Images courtesy of http://www.fs.fed.us -Forest Service aerial survey data: Digital Aerial Sketchmapping System Survey acreage, insect agent, damage severity, etc. - Advantages: GPS tracking (easier) Multiple base maps produce greater accuracy GIS and GPS ready Fast and can cover a large area/widely available - Disadvantages: Error and inaccuracy

13. 2002 Landsat TM 5 image and FS aerial survey dataset 2002 - 78 total polygons used -Normalized Difference Vegetation Index (NDVI) and 3 rd Tasseled Cap (wetness) performed upon 2002 image

14. Elevation and aspect - Elevation affects temperature: One study showed that brood production in a given thickness of bark was inversely related to elevation [3] A study in the Snowy Range showed 2,566 m as transition point between patchy lower forest (montane) and solid lodgepole pine (subalpine) [4] - Aspect serves as a proxy for amount of insolation: Aspect partially controls plant species distribution, spring thaw intensity, soil and air temperature, etc. Four cardinal directions (North, East, etc.) used for slope aspects in this analysis - Mean elevation and the majority aspect calculated for each polygon - Resulting set of wetness values stratified by polygon area, elevation class, and slope aspect:

15. Results

21. Discussion - In the Medicine Bow National Forest, are beetle impacted sites exhibiting a different reflectance pattern than non-impacted areas? Statistically significant decrease in wetness and NDVI - Does the rate of decline vary according to different elevation zones? Upper elevation class exhibited lower wetness values than the elevation class below 2567 m Upper class demonstrated lower NDVI values than the elevation class below 2567 m Function of fewer trees being affected or fewer trees in general? - Does damage vary according various slope aspects? Lower NDVI values on south and east facing slopes (somewhat expected) Driest slope on average was the east facing, south the wettest (?): Anomalous effect caused by precipitation?

22. For Future Research -Test other variables of interest (stand density, crown cover, species type): - Dependent on datasets’ detail and age, i.e. RIS - Larger sample sizes needed: - To overcome uncertainty w/ pixel size and shapefile boundaries - To more soundly determine whether statistical relationships exist between these environmental variables and others not yet tested - Larger temporal scale (more years) - Management implications: Use Landsat imagery 1 st : Save resources Know where to direct aerial surveying Update aerial survey tech. (integrated GPS and video)

23. In conclusion… - A combination of Landsat TM imagery transformations guided by FS aerial survey data was able to demonstrate tree mortality in the Sierra Madre Range: Landsat 1 st, aerial surveying 2nd - Additional management units will have to be analyzed to more soundly determine whether varying environmental factors influence stand mortality rates Strengthen statistical relationships - If such relationships can be determined, they could serve in producing valuable map products for land managers to use Especially in unaffected areas!!!

24. [1] Wyoming: Beetle Epidemic. Retrieved from http://www.fs.fed.us/r2/bark-beetle/fact- sheets/CumulativeAcresNFs-Counties.pdf [2] Digital Aerial Sketchmapping. Retrieved from http://www.fs.fed.us/r2/resources/fhm/aerialsurvey/aerial_survey/FS_dasm.pdf [3] Amman, G.D. 1977. The role of the mountain pine beetle in lodgepole pine ecosystems: impact on succession. Mattson, W.J. (ed). Spring –Verlag, [4] Billings, W.D. 1990. The mountain forests of North America and their environments. Pp. 47-86 In Plant biology of the basin and range. Osmond, C.B., Pitelka, L.F., and G.M. Hidy. (ed). Springer Verlag, 375 pp. Photo from July 2008, by author