Modeling tree height and basal area in the Finger Lakes National Forest, NY

There was definitely a lot of scatter going on, but it was good to visualize how well things were lining up in the model versus what was actually collected in the field.The plot I made for tree height (it looks better than the basal area one did so I’m using it).The code I used to generate my modeled vs. actual height plot.Lastly, I went in and applied my models of tree height and basal area to the entire Finger Lakes National Forest! So, I called in the big, fat “grid.csv” file with the UTM coordinates and grid metrics data for the whole forest and then used my models generated before to predict height and basal area values for every single 15 x 15 m plot in the study area. Then I saved this data in a CSV file and YAY! I was done.My code for making my height model for the entire study area. I used the head function a few times just to see how things were going before I generated another massive CSV file.Some patterns and thoughts I had after being able to view my models:Below are some maps made in arcGIS of what a small area of the forest looked like, if you were curious! I wanted to investigate what tree height and basal area values were all around the Finger Lakes National Forest and by displaying the predicted values I got in R I was able to do that which was really cool to look at actually.a) is satellite imagery, b) shows how old the forest is in the area, c) shows the predicted tree height values per cell and d) shows basal area.Generally speaking, tree heights didn’t end up too far over 25 m while basal area hovered at values under 2 m²/ha..Overall, the models appeared to match up pretty well with what we see for tall/old looking trees in satellite imagery and also it looked like height and basal area were sort of positively correlated with each other AND forest age — I ended up looking into how basal area and height varied with forest age too since there are parts of the Finger Lakes National Forest that have been forested longer than other parts.If I had some more time I think I would look more into whether or not height and basal area are actually related to forest age..I would also have given some more thought to which variables I used and how many I used in order to try and get some more of the variation in the data explained, especially for basal area since only a little over a quarter of the variation we saw was explained by the model.References:Brubaker, K..M., Q..Johnson, and M.WK.Kaye..(2018)..“Spatial patterns of tree and shrub biomass in a deciduous forest using leaf-off and leaf-on lidar.” Canadian Journal of Forest Research..48(9): 1020–1033.Murphy, P..(2018)..Unpublished data… More details

Leave a Reply