Friday, April 30, 2010

Start of Rainier GIS Analysis

I began my GIS analysis of Rainier the past week or so. I had thought I could do it pretty quick but I forgot that there are 28 glaciers I’m doing analysis for, over three years and for each year for a different outline (Nylen 94’ and 1913). Thus, I still have a bunch to do but have made good progress so far. First, there were a couple occasions where in one of the outlines (94 or 1913) two glaciers were combined into one outline. For example, the Nisqually and the Wilson were combined in the 13 outline while they were separate in the 94 outline. Thus I used the 94 outline split line along with the hillshade to split the two for 1913. There were several other instances of this and splitting them took a bit of time as I had to be very careful not to skew the data. There were also a couple occasions where the outline was a little bit off of the glacier which I fixed and made note of. Also, something neat that I noticed... In the 1913 outline the Cowlitz and Ingraham were combined and the 1994 Nylen outline the Ingraham went into the Cowlitz... but the Glims database online shows that they remain separate to their ends. I took a closer look and with the high resolution LiDAR hillshade they are indeed two seperate glaciers to their distinct toes so I followed along the ridge using the hillshade and split them that way (similar to what Glims shows). I’m not sure if it was clear before the LiDAR came out how distinctly separate they are...

So, as of now I’ve split both the 94 and 1913 outlines into separate glacier outlines for the 28 glaciers I’ll analyze. I’ve gotten through all 28 for the 1987 band ratio using the 1994 Nylen outline as the clip as the colorful image on the right shows. I’m then going to use the same color scheme as I did for Hood to be consistent. I quickly did a couple from 1992 (the brown color) and 2005 (the blue) for the Emmons and Winthrop Glaciers (lower right image).

Because Rainier is so large it has taken me all day to get the slope file layer to calculate in ArcCatalog so I’ll get to the slope and true area calculations next week as well as finishing up 92 and 2005 for the 94 outline and all three years for the 1913 outline.

From the looks of it I think the 1913 outline is going to work better because it gives a little more freedom and because the 1994 outline was created 7 years after my 1987 Landsat ratio images (although it was based on glacier extent, not debris free ice, so it may still be valid).

I wasn’t able to get to Melinda about the conference in Portland as I got way too caught up in fixing outlines and getting myself ready to churn out the rest next week. The abstract is due May 10th so I’ll make sure to get that all done next week.

Let’s see... 28x2 = 56 more for the 94 outline and 28x3 = 84 for the 1913 outline, for a total of 140 more clips and display changes, and then over 150 slope zonal statistics to get the true areas... looks like I’ll be busy the next week or so. I’ll try to get an outline to you as soon as I get through this processing.

Cheers - Jon

No comments: