The Sawtooth National Wilderness
- The Gory Details -
For those interested in the process/criteria I used to produce the Sawtooth National Wilderness Area lake count, read on...
DISCLAIMER: I'm not specifically trained in any field that would qualify me as an expert at producing an authoritative report on the number of water-bodies in the geographic area. I'm not at all sure what the formal definition of a "lake" is. Regardless, the methods I used to gather the data provided are described below.
DATA SOURCES: The primary source of information were the aerial images provided by GoogleEarth. Not just the current image but all available historical imagery provided by GoogleEarth (11 image-sets from 8/27/1998 to 7/20/2013). I also used the 1984 USGS Topological maps, images placed by users on GoogleEarth and in a few cases, actual visits to the lake locations.
SELECTION CRITERIA: Not long into this project I started using the term "water-body" in addition to "lake", just to relieve myself of the possibility of incorrectly using the term "lake". I use them interchangeably however. The primary reason for this was my desire to include rather small bodies of water. I did not have a formal size criteria. I basically used the viewable resolution of the GoogleEarth aerial images to identify as small of water-bodies as I could. Roughly, this resolved to water-bodies as small as about 50 feet across. Criteria I chose to ignore in disqualifying a water-body were: size, seasonal persistence, water depth, altitude, shape and ice cover.
Several water-bodies in my list are formed each year by ice dams, which melt away each spring and lets all or most of the water escape. Yet others, in some years, are never exposed from beneath a glacier that in any particular year may not melt off. In short my selection criteria was quite broad. If the water-body was big enough to see in aerial images, was within the official boundary of the Sawtooth National Wilderness Area and held water at least some point during the year, I included it.
TOOLS: Because the project spanned a 2-year period, several of the tools I was using change revision. I completed the project for first web publishing using: GoogleEarth v7.1.2.2041, MS Excel 2007, kmlcsv Converter v2.2.1, MS Expression Web v4
SEARCH PROCESS: The searching process was brute-force, I just methodically scanned the aerial images (GoogleEarth) visually, north to south, west to east placing pushpins in water-bodies as they were found. Pushpin placement was located as close to the geometric center of the water-body as visually possible. I cross referenced finds of larger water-bodies against the USGS Topo maps to confirm and look for official lake names. The USGS Topos have far fewer records of lakes than are visible in GoogleEarth. The USGS Topos were also scanned and lake locations verified (or not) in GoogleEarth. I say "or not" because there were a handful of lakes recorded in the USGS Topos that just simply don't exist. I repeated this process with several of the highest resolution images available in GoogleEarth.
Later when higher resolution images became available this process was repeated again. Lastly, I use a random search method of just relooking more closely in areas geographically more likely to contain bodies of water. Over this process additional subtle water-bodies were found and some false positives were removed. Most of the false positives ended up being meadows that looked like bodies of water in poorer resolution aerial images.
Older aerial images proved most valuable at finding water-bodies "hiding" in more recent imagery. Examples are water-bodies hiding beneath temporary glaciers that may not have melted off that year or hadn't at the time of the image, and, in several locations the shadows of high mountain peaks obscured certain lakes from view in some images but not in others.
NAMING PROCESS: The majority of water-bodies in the Sawtooth National Wilderness Area are unnamed, about 90%. I needed at least a temporary name. I decided to just number them. I chose to number all the lakes, named or not. The unnamed ones I just call "Unnamed - #". The number of each lake is meaningless other than it creates a unique name for each lake and they are sorted from north to south, "Unnamed - 1" being the furthest north. The northerly ranking is relative to the location of the geometrically centered (eye-balled) pushpin. This simply means that larger lakes can have northern shore boundaries further north than another lake(s) with a lower naming number.
Some of the named lakes in the Sawtooth National Wilderness Area are named as clusters of lakes. For example, there is a cluster of 6-7 lakes called the Trail Creek Lakes. In these cases, and because it wasn't always clear which lakes were included in the cluster, I only gave the largest of the lakes the official name and named the remainder "Unnamed - #". I admittedly gave this little thought. There's likely a better way of handling this but for now, it is what it is.
Over the course of the project I worked with several versions of GoogleEarth aerial images. I noticed that they aren't all registered exactly the same. I didn't look into this but common sense would say that registration accuracy is a subject unto itself and that the complications of registration I'm leaving to GoogleEarth. I mention it because I note that as Google tweaks their images the water-bodies shift relative to the pushpins. So if you see a pushpin not in the geometric center of a body of water, that's likely the reason.
FILE GENERATION: I keep my master list of water-bodies in a Places folder in GoogleEarth. An exact copy of it is available on the Download page of this site. When updating the list (adding or removing pushpins) I write out the list to a .kmz file. Using a convertor I convert it to a .csv file. Using Excel I sort the list based on the latitude. I use the feature of Excel to maintain the name number. After the list is sorted it is converted back to a .kml file and loaded back into GoogleEarth and the revision number incremented. This website is then updated with the latest version.
Note that this methodology has the unpleasant side-effect of possibly changing a water-bodies northerly number from version to version. Again, I didn't put a lot of thought into this, there might be a better way. I used it because I prefer that the number be a nice orderly north-to-south ordering of the water-bodies.
Email me with questions, comments, suggestions: cal @ ckmcdonald . com