Check out this motherlode!
Hello everyone! One week in and we are 14% funded with 26 backers- thank you all so much. Remember, each backer wins us a portion of the challenge grant from Robert Downey Jr.'s Footprint Coalition, so even a $1 donation helps! We have drawings coming up on January 27th (An 11x14" aluminum print ($1 min or 1 referral) and February 12th (Another aluminum print ($1 min or 1 referral) and a 5-night stay ($25 min or 10 referrals) outside of Glacier National Park!) , so please stay tuned and help us get the word out.
Let's start walking through the methods driving this project and why our results are so important to wildlife in the northern Rocky Mountains. In the process, we'll even share the location of a very large patch of huckleberries!
First things first- we need to know what to teach the computer to look for. Some vegetation is a unique color at a certain time of year, some has different phenology (starts growing early in the year, is an evergreen), and some vegetation has physical structure that makes it stand out. If we give the computer a large number of a wide variety of examples of a focal plant, we can train it pretty well to pick these plants out of an enormous landscape. Here's an example from Whitefish Mountain:
Now, I'm sure most of us can pick out the color change between the July image on top and the early September image on bottom- huckleberry shrubs! Do you think bears know where this meadow is? Definitely! The mountain is actually a ski resort that sends followers updates on where to pick, and warn pickers to bring bear spray https://skiwhitefish.com/huckreport/ , so I'd say that's an emphatic yes!
Join us next time for an update on common juniper mapping- can you guess what characteristics we'll try to focus on in our attempt to map them?
0 comments