A Common Language for Computer Programs

I spent a few days this week preparing to move our modeling into the more complicated spatiotemporal domain, which was a fun learning experience. One of the biggest features that our spatiotemporal model needs to incorporate is the complex spatial features of our model domain. Since we know that many of our organisms have preferred habitat types, and some of these habitat types are physical features rather than temperature or salinity ranges, we need to include these physical features in the model to help inform habitat suitability and survivorship calculations. This week, therefore, I learned how to take our maps from ArcGIS and put them into our modeling software by creating ASCII files. An ASCII file is a standardized file format (like a .txt or .csv file), that stores information as text, which can be read by many computer programs. If you're not familiar with computing or computer files, you can think of an ASCII file as a common language between computer programs. By generating an ASCII file, I can take data from one computer program and give it to another program to understand. Such is the case here, where I'm taking maps from a geospatial mapping and analysis program (ArcGIS) and reading them into our ecosystem modeling program. The ASCII file contains rows and columns of information, and because our maps have exact coordinates for the physical features, we need to ensure that we are assigning the data to the correct row and column combination. Here, our amazing graduate student, David, wrote a set of code that evaluates the percent overlap between our map features and the grid cells, so that each row and column combination fits together like a mosaic and paints a beautiful picture of our marshland and oyster reef habitats. 

What I especially appreciate about the process of converting between the ArcGIS map and the ASCII file is that I can import the completed file into our modeling software to visualize the data and then import that same file into the mapping software, which provides a higher resolution image (clearer, more detailed picture) of the same map. Therefore, I get to check the outcome multiple times to evaluate whether or not the conversion and coding were successful.

Next week is going to be a break from this geospatial modeling work and back to oyster mortality modeling, as I have some exciting news to share about the other research that I'm doing. It'll have to wait until next week though, so stay tuned!