Machine learning techniques, in concert with an exponentially growing availability of data, have the potential to enable game-changing innovation in how we think about and approach environmental, infrastructure, and societal challenges. Leveraging these tools in a way which ensures results are trustworthy is critical, particularly given taxpayer funding and effects of decisions on communities. However, the most pressing current challenge facing our discipline is the lack of meaningful ML environmental studies (work is being done but frequently with only shallow knowledge of the ML pipeline or of the environmental significance and implications). The workshop will work toward a framework for asking the right questions to develop well-rounded studies leveraging ML tools. Organizers: Amy Mueller, Northeastern University; Branko Kerkez, University of Michigan; Kate Newhart, West Point; Huichun Zhang, Case Western University