Clipping Rasters with GeoJSON Polygons

A recent side project we’ve been working on at SNAP requires clipping data from a user selected polygon.  Plenty of people are already using the gdalwarp utilities for data clipping, but using GeoJSON seems to be a bit less popular.  In our case, since the polygon is being passed over the web, JSON seemed like a natural format.  And since the GeoJSON specification is designed specifically for the job, it seemed like a natural solution.

Here is the area we want to clip:



Technical position recruitment

In my opinion, if you are applying for a technical position, where it is often hard to convey skill and knowledge in a resume aside from ridiculously long lists of software, show me your skills.  I’m thinking about requiring all applicants to provide “examples of work”.  I don’t think a portfolio of work should be limited to the traditional web or printed design fields.  If  your strengths are in programming, you should have a fairly well developed set of code to share.  Did that code contribute to a program?  Show me it.  If your skills are in cartography (which is half artistic skills), then show me the maps.  Web design skills, show me the sites and exactly what sections you contributed to.  Put it on the web, just show me.  If you are unable to produce any products or examples of work, then you likely need to think twice about your job search strategy.  Maybe it’s because I’ve always been a visual learner, but I like to see things.

In addition, learn how to find information.  Yes, collaborating with colleagues is very important and everyone gains from continuous interoffice communication, skill transfer, etc, etc, but the organization will not advance if many people stagnate in their quest for knowledge.  Google it, read, explore, stretch your mind.  Constructive criticism is important, but new ideas are usually better.  Go find them, then steal them, develop them further, then share them back.  Tell me how you do this on a daily basis, as it’s very valuable to an employer.

Colors in charts

more emails from Bruce…

I came across this reference, which contains a number of good points regarding the use of colors in charts:

Happily, many years ago I learned Tufte and his approach to visual design of communicating quantitative information:

Life changing.
In particular, this book looks to be relevant for SNAP

and from Kristin…
This is another useful site with some info about schemes:

Coding best practices for scientists

An email Bruce sent out…
Hi — over the past few months I’ve been aggregating some links to articles and papers that recognize that the goals of scientists who write code are different from “industry” programmers, but there’s some tools and practices that are common in the industry that could be of enormous utility for scientists who create programs.  This post has some links and references, in case you are interested.
First, a general scholarly overview of software development practices that are useful for scientists:
Next, an interactive introduction to using Github (source code control):
Here’s an interesting resource that has general resources and tutorials for other software development practices, such as automating tasks:
Finally, a few links to some narrative articles about why some of these practices are important:


Welcome to the Scenarios Network for Arctic and Alaska Planning (SNAP) blog!  We’ll use this to publish information about how we do our work, on topics including R, data processing and visualization, climate model software development, and data processing infrastructure.