Tree Height Growth in Petawawa Research Forest

Monitor through LiDAR CHMs from 2005 to 2018

Zhongfang Qian

Context

Monitoring the growth of trees is important for sustainable forest management. The traditional method of monitoring forest growth at a broad level is timber cruising by humans, while modern remote sensing technology, especially Lidar, is usually used for monitoring at the single-tree level or stand-level. This project uses Light Detection and Ranging (LiDAR, i.e. airborne laser scanning (ALS)) data from 2005, 2012, 2016, and 2018 and forest survey data from 2007 from the Petawawa research forest (PRF) supersite to monitor the tree height growth through a time series and build relationships between stand ages and dominant tree heights (DH). 

Study Area

The PRF was established in 1918 to create and transfer knowledge that could be used to improve the quality, productivity, and health of Canada's forests, and is the oldest sustainably operated research forest in Canada. Over the 100-year history of the PRF, many ground sample plots and research experiments have been established. The data holdings, long-term infrastructure, and ease of access have made this data source an important one. More than 1900 data records, including multiple airborne laser scan datasets and associated derived data (i.e., digital terrain models, canopy height models), airborne imagery, and ground patch data, are publicly available for download from the Canadian National Forest Information System (CNIS) or its supersite. 

Method

Field data processing

By overlaying and intersecting every outline and 2007 inventory range, landcover classification based on stand age can be done on the areas which contains all types of data.

LiDAR data processing

The areas are devided into 10x10m grids and the dominant tree height of each grid can be derived. Based on variation of height in the time series and normal tree ranges for the native species, the whole forest can be classified as healthy/disturbed and misclassified areas.

Statistical analysis

Different linear models are developed and tested for the healthy data in each landcover type.

Results

  • 61.03% of forest is healthy and only 0.97% is misclassified
  • Most of healthy and disturbed areas is Mid-Seral forest, while most of misclassified areas is Early Seral forest
  • Same as the overall area composition, the least of healthy, disturbed and misclassified forest is Old Growth forest
  • some concentrated areas which may be clear-cut areas
  • The maximum growth rate happened from 2005 to 2012 in the Early-seral forest, which is 182.16%/year
  • unusual growth of growth rate from 2012 to 2018

PRF-landcover classification

multiple linear models for each landcover type

multiple linear models for the whole forest

Discussion

This method represents a cost and time-efficient alternative to manually measuring the height of dominant trees in closed forests. 

Results analysis

  • small percentage of disturbed areas is caused by the random measurement error (e.g. detection of birds) that happened in a single place in a single year
  • the relationship between the stand ages and the DHs can be described by the max/mean/median/min values of DHs in the stands of the same age
  • these models are very applicable and can be used to predict DHs
  • the process of building the model is easy to apply in different forests

Management suggestions

  • potential misclassification areas should be re-inventoried by humans or referred to in the previous logs
  • management approaches (e.g. thinning) in the areas where restrict the tree's healthy growth
  • areas other than Old Growth with a low growth rate (<0.01%) should also be paid attention to

Limitations and future direction

  • only focuses on tree height growth
  • develop specific models like volume prediction or biomass estimation with sample plots in each land cover type
  • the results of this paper depend in large part on the accuracy of the inventory map
  • Deep learning can be used in forest cover classification
  • Digital aerial photogrammetry (DAP) can also be combined with Lidar data

Alexander, C., Korstjens, A. H., & Hill, R. A. (2018). Influence of micro-topography and crown characteristics on tree height estimations in tropical forests based on LiDAR canopy height models. International Journal of Applied Earth Observation and Geoinformation, 65(October 2018), 105–113. https://doi.org/10.1016/j.jag.2017.10.009

BC Forestry. (2014). Appendix D – Codelists: Age Class Code. Retrieved from: https://www.for.gov.bc.ca/hfp/publications/00026/fs708-14-appendix_d.htm#ad_03

Benediktsson, J. A., Swain, P. H., & Ersoy, O. K. (1990). Neural network approaches versus statistical methods in classification of multisource remote sensing data. IEEE Transactions on Geoscience and Remote Sensing, 28(4), 540-552. https://doi.org/10.1109/TGRS.1990.572944

Buffington, K. J., Dugger, B. D., Thorne, K. M., & Takekawa, J. Y. (2016). Statistical correction of lidar-derived digital elevation models with multispectral airborne imagery in tidal marshes. Remote Sensing of Environment, 186, 616-625. https://doi.org/10.1016/j.rse.2016.09.020

Davis, G., British Columbia. Forest Service, & HathiTrust ETAS Collection. (2004). A modified timber cruise for the inventory of dead wood in coastal forests: A field trial. British Columbia Forest Service, Vancouver Forest Region.

Duncanson, L., & Dubayah, R. (2018). Monitoring individual tree‐based change with airborne lidar. Ecology and Evolution, 8(10), 5079-5089. doi:http://dx.doi.org.ezproxy.library.ubc.ca/10.1002/ece3.4075

Fradette, M. S., Leboeuf, A., Riopel, M., & Bégin, J. (2019). Method to reduce the bias on digital terrain model and canopy height model from LiDAR data. Remote Sensing, 11(7). https://doi.org/10.3390/RS11070863

Gaveau, D.L., and Hill, R.A. (2003). Quantifying canopy height underestimation by laser pulse penetration in small-footprint airborne laser scanning data. Canadian Journal of Remote Sensing, 29, 650–657.

Gislason, P. O., Benediktsson, J. A., & Sveinsson, J. R. (2006). Random forests for land cover classification. Pattern Recognition Letters, 27(4), 294-300. https://doi.org/10.1016/j.patrec.2005.08.011

Hayat, A., Hacket-Pain, A. J., Pretzsch, H., Rademacher, T. T., & Friend, A. D. (2017). Modeling tree growth taking into account carbon source and sink limitations. Frontiers in Plant Science, 8, 182-182. https://doi.org/10.3389/fpls.2017.00182

Jankiewicz, L. S. (1972). A cybernetic model of growth correlation in young apple trees. Biologia Plantarum, 14(1), 52-61. https://doi.org/10.1007/BF02920902

Khosravipour, A., Skidmore, A.K., Isenburg, M., Wang, T., and Hussin, Y.A. (2014). Generating pit-free canopy height models from airborne LiDAR. Photogrammetric Engineering & Remote Sensing, 80, 863–872.

Kim, E., Lee, W. K., Yoon, M., Lee, J. Y., Son, Y., & Salim, K. A. (2016). Estimation of voxel-based above-ground biomass using airborne LiDAR data in an intact tropical rain forest, Brunei. Forests, 7(11), 27–36. https://doi.org/10.3390/f7110259

Koch, G. W., Jennings, G. M., Sillett, S. C., & Davis, S. D. (2004). The limits to tree height. Nature, 428(6985), 851-854. https://doi.org/10.1038/nature02417

Kvålseth, T.O. (1985). Cautionary note about R2. American Statistician, 279–285.

Leckie, D.G. (1990). Advances in remote sensing technologies for forest surveys and management. Canadian Journal of Forest Research, 20, 464–483.

Lefsky, M. A., Cohen, W. B., Parker, G. G., & Harding, D. J. (2002). LiDAR remote sensing for ecosystem studies. BioScience, 52(1), 19–30. https://doi.org/10.1641/0006-3568(2002)052[0019:LRSFES]2.0.CO;2

Lo, C. S., & Lin, C. (2013). Growth-competition-based stem diameter and volume modeling for tree-level forest inventory using airborne LiDAR data. IEEE Transactions on Geoscience and Remote Sensing, 51(4), 2216–2226. https://doi.org/10.1109/TGRS.2012.2211023

Maltamo, M., Eerikäinen, K., Packalén, P., & Hyyppä, J. (2006). Estimation of stem volume using laser scanning-based canopy height metrics. Forestry, 79(2), 217–229. https://doi.org/10.1093/forestry/cpl007

Næsset, E. (2002). Predicting forest stand characteristics with airborne scanning laser using a practical two-stage procedure and field data. Remote Sensing of Environment, 80(1), 88–99. https://doi.org/10.1016/S0034-4257(01)00290-5

Nigh, G. (2017). The effect of leader damage on white spruce (picea glauca) site tree height growth and site index. Canadian Journal of Forest Research, 47(9), 1153-1160. https://doi.org/10.1139/cjfr-2017-0056

Nilsson, M. (1996). Estimation of tree heights and stand volume using an airborne lidar system. Remote Sensing of Environment, 56, 1 – 7.

Petawawa Research Forest Supersite. (2005). Airborne LiDAR 2005 LAS files [Data file]. Retrieved from: https://ln.sync.com/dl/1afc42aa0/x3qnwycs-k9mxewce-e4phit4q-esetiq3u?sync_id=1164390568008

Petawawa Research Forest Supersite. (2007). PRF forest inventory 2007 [Data file]. Retrieved from: https://opendata.nfis.org/downloads/petawawa/Vector/Forest%20Inventory/prf_forest_inventory_2007.zip

Petawawa Research Forest Supersite. (2012). Airborne LiDAR 2012 LAS files [Data file]. Retrieved from: https://ln.sync.com/dl/1afc42aa0/x3qnwycs-k9mxewce-e4phit4q-esetiq3u?sync_id=1164390568008

Petawawa Research Forest Supersite. (2016). Airborne LiDAR 2016 LAS files [Data file]. Retrieved from: https://ln.sync.com/dl/1afc42aa0/x3qnwycs-k9mxewce-e4phit4q-esetiq3u?sync_id=1164390568008

Petawawa Research Forest Supersite. (2018). Airborne LiDAR 2018 LAS files [Data file]. Retrieved from: https://ln.sync.com/dl/1afc42aa0/x3qnwycs-k9mxewce-e4phit4q-esetiq3u?sync_id=1164390568008

Place, I.C.M. (2002). 75 years of research in the woods: A history of Petawawa Forest Experiment Station and Petawawa National Forestry Institute, 1918 to 1993. General Store Publishing House, Burnstown, Ontario.

Ryan, M.G. (USDA, Forest Service Rocky Mountain Experiment Station, Fort Collins, CO.), & Yoder, B. J. (1997). Hydraulic limits to tree height and tree growth. Bioscience, 47(4), 235-242. https://doi.org/10.2307/1313077

Sánchez A. J., Oliver, P., Estornell, J., & Dopazo, C. (2018). Estimación De Variables Forestales De Pinus Sylvestris L. En El Contexto De Un Inventario Forestal Aplicando Tecnología LiDAR Aeroportada (ESTIMATION OF FOREST VARIABLES OF PINUS SYLVESTRIS L. IN THE CONTEXT OF A FOREST INVENTORY APPLYING AIRBORNED LIDAR TECHNOLOGY). GeoFocus Revista Internacional de Ciencia y Tecnología de La Información Geográfica, 79–99. https://doi.org/10.21138/gf.509

Shamsoddini, A., Turner, R., and Trinder, J.C. (2013). Improving LiDAR-based forest structure mapping with crown-level pit removal. Journal of Spatial Science, 58, 29–51.

Shaw, J. W., & Biodiversity Heritage Library. (1910). How to cruise timber, adapted for experienced cruisers, loggers, foresters, claimants, or for any one desiring to learn to estimate timber. Publisher not identified.

Silva, C. A., Hudak, A. T., Vierling, L. A., Loudermilk, E. L., O’Brien, J. J., Hiers, J. K., Jack, S. B., Gonzalez-Benecke, C., Lee, H., Falkowski, M. J., & Khosravipour, A. (2016). Imputation of Individual Longleaf Pine (Pinus palustris Mill.) Tree Attributes from Field and LiDAR Data. Canadian Journal of Remote Sensing, 42(5), 554–573. https://doi.org/10.1080/07038992.2016.1196582

Solvin, T. M., Puliti, S., & Steffenrem, A. (2020). Use of UAV photogrammetric data in forest genetic trials: measuring tree height, growth, and phenology in Norway spruce (Picea abies L. Karst.). Scandinavian Journal of Forest Research, 35(7), 322–333. https://doi.org/10.1080/02827581.2020.1806350

White, J. C., Chen, H., Woods, M. E., Low, B., & Nasonova, S. (2019). The Petawawa research forest: Establishment of a remote sensing supersite. Forestry Chronicle, 95(3), 149–156. https://doi.org/10.5558/tfc2019-024

White, J., Stepper, C., Tompalski, P., Coops, N., & Wulder, M. (2015). Comparing ALS and image-based point cloud metrics and modelled forest inventory attributes in a complex coastal forest environment. Forests, 6(12), 3704-3732. https://doi.org/10.3390/f6103704

White, J. C., Tompalski, P., Vastaranta, M., Saarinen, N., & Stepper, C. (2017). A model development and application guide for generating an enhanced forest inventory using airborne laser scanning data and an area-based approach.

White, J. C., Wulder, M. A., Varhola, A., Vastaranta, M., Coops, N. C., Cook, B. D., Pitt, D., & Woods, M. (2013). A best practices guide for generating forest inventory attributes from airborne laser scanning data using an area-based approach. In Forestry Chronicle (Vol. 89, Issue 6). https://doi.org/10.5558/tfc2013-132

Yuan, Q., Shen, H., Li, T., Li, Z., Li, S., Jiang, Y., Xu, H., Tan, W., Yang, Q., Wang, J., Gao, J., & Zhang, L. (2020). Deep learning in environmental remote sensing: Achievements and challenges. Remote Sensing of Environment, 241(January), 111716. https://doi.org/10.1016/j.rse.2020.111716

Zhang, C., Sargent, I., Pan, X., Li, H., Gardiner, A., Hare, J., & Atkinson, P. M. (2019). Joint Deep Learning for land cover and land use classification. Remote Sensing of Environment, 221, 173–187. https://doi.org/10.1016/j.rse.2018.11.014

Zhao, M., Wang, T., Wu, H., Nie, D., Li, M., Liu, J., Xie, M., Li, S., & Zhuang, B. (2021). A new method for estimating the noise scale factor (NSF) and random errors in lidar observations. Applied Physics. B, Lasers and Optics, 127(2)https://doi.org/10.1007/s00340-021-07572-2

multiple linear models for each landcover type

multiple linear models for the whole forest