An Extremely Productive Day-Probably My Best Post

June 19, 2013

While not visiting the field today, the team met up at 8 am this morning to hash out research topic ideas and offer alternative ideas and critical construction. I presented my idea of calculating Biomass from Photoscan on a micro site level and everyone seemed to like it. After some discussion and thought, it seemed the best area for study was the Hao and Monkey Pod Trees located directly at the staging area (shouldn’t get lonely working on my project). However, discussion arose between Dr. Lee and I when considering how to calculate Biomass in a traditional manner without a preexisting  models. I did some infant stage research on this topic and thought about about it pretty much the whole day. This is what I came up with….

To calculate biomass from remote sensing, I will need some estimate of biomass from the field, and remotely sensed data from the same locations. I then plot one against the other and measure the statistical relationship between them. By convention, the thing you want to predict should be on the Y axis, and is called the dependent variable. The X axis would be the remotely sensed data – NDVI in my case. The kind of analysis I will need is called a linear regression, and it amounts to finding the line that goes through the data and predicts Y given X. This is just the familiar Y = mX + b.  Once I know m and b, I can predict Y at locations where it has not been measured but I know X.
 —–Obviously this is fairly simplistic and basic, but it plays an important role in developing a theoretical model for my project.
I contacted a professor of mine who works on calculating biomass with Dr. David Clark and he pointed me to some literature which might lay out a useful plan for my research project. He suggested that the paradigm for calculating Biomass is in need of change and that real hard, ground “truthing” is needed to “accurately” estimate the biomass of a given area unit.
After a discussion with Paul and Dr. Wechsler in the later afternoon, Lidar imagery does exist to the point of my study area, In my opinion, this completely changes the game and I will need tomorrow to know exactly how useful the Lidar imagery will be for me and the project in the coming week.
Cole Anwyl Walters
University of Maryland, College Park