There is a common theme at play in our talk of data generally, of digital earth data more specifically, and of environmental monitoring most specifically: more data leads to more action and, ultimately, to societal good. This data-to-action framework is troubled. Its taken-for-grantedness prevents us from attending to the processes between data and action. It also dampens our drive to investigate the contexts of that data, that action, and that envisioned societal good. In this dissertation, I deconstruct this data-to-action model in the context of Landsat, the United States' first natural resource management satellite. First, I talk about the ways in which Landsat's data and instrumentation hold conflicting narratives and values within them. Therefore, Landsat data does not automatically or easily yield action toward environmental preservation, or toward any unified societal good. Furthermore, I point out a parallel dynamic in STS, where critique is somewhat analogous to data. We want our critiques to yield action, and to guide us toward a more just technoscience. However, critiques—like data—require intentional, reconstructive interventions toward change. Here is an opportunity for a diffractive intervention: one in which we read STS and remote sensing through each other, to create space for interdisciplinary dialogue around environmental preservation. A focus on this shared goal, I argue, is imperative. At stake are issues of environmental degradation, dwindling resources, and climate change. I conclude with beginnings rather than endings: with suggestions for how we might begin to create infrastructure that attends to that forgotten space between data, critique, action, and change. / Doctor of Philosophy / I have identified a problem I call the data-to-action paradigm. When we scroll around on Facebook and find articles –– citing pages and pages of statistics –– on our rapidly melting glaciers and increasingly unpredictable weather patterns, we are existing within this paradigm. We have been offered evidence of looming, catastrophic change, but no suggestions on what to do about it. This is not only happening with climatological data and large-scale environmental systems modelling. Rather, this is a general problem across the field of Earth Remote Sensing. The origins of this data-to-action paradigm, I argue, can be found in old and new rhetoric about Landsat, the United States’ first natural resource management satellite. This rhetoric often says that Landsat — and other natural resource management satellites’ — data is a way toward societal good. The more data we have, the more good will proliferate in the world. However, we haven’t been specific about what that good might look like, and what kinds of actions we might take toward that good using this data. This is because, I argue, Earth systems science is politically complicated, with many different conceptions of societal good. In order to be more specific about how we might use this data toward some kind of good we must (1) explore the history of environmental data, and figure out where this rhetoric comes from (which I I do in this dissertation), and (2) encourage interdisciplinary collaborations between Earth Remote Sensing scientists, social scientists, and humanists, to more specifically flesh out connections between digital Earth data, its analyses, and subsequent civic action on such data.