Andrew Slater Science Tribute
Andrew (Drew) Slater, land modeler extraordinaire, passed away in the beginning of September 2016. Drew was 44.
Drew was a force within the land modeling community. He transformed our capabilities to simulate terrestrial processes and change in cold regions. He was best known for his noteworthy advances in modeling snow, permafrost, and hydrologic processes in Arctic and mountain regions. A complete scientist, Drew was a theorist, a modeler, an experimentalist, and a teacher, equally competent writing model code, analyzing satellite data, stomping through snow to collect measurements, and explaining the impact of climate change on skiing to school kids. Drew pursued science and life with passion – he had keen insights and intellect; he sought innovative, meaningful, and thorough solutions to earth science problems; and he was perpetually providing thoughtful critiques of ongoing work and engaging in fierce debates on ways to accelerate advances in land modeling capabilities. Drew had a penchant for excellence – he thought that the research system was broken, and he strived to fix it. In this Tribute we wish to celebrate Drew’s scientific accomplishments and the enormous impact that he had on our field and our lives.
While in graduate school Drew focused his attention on developing and evaluating snow parameterizations for use in climate models. In this work Drew led the snow model inter-comparisons in PILPS (the Project for the Inter-comparison of Land surface Parameterization Schemes). This work was vital for the land modeling community to understand and address weaknesses in snow parameterizations. In Drew’s PILPS paper he did not just document differences among models; critically, Drew was able to identify specific reasons for wildly different model behavior. Drew’s insights advanced understanding on how model representations of sub-grid heterogeneity affect the amount of energy available for ablation, and how inter-model differences in sensitivity to downward longwave radiation can be explained by stability-induced feedbacks. Many of us today still go back and re-read Drew’s early PILPS paper to try and understand the behavior of modern land models.
After completing his PhD in 2003, Drew focused a large part of his efforts to advance capabilities for streamflow forecasting in snowmelt-dominated river basins. Drew’s most noteworthy contribution in this area was his work on ensemble snow data assimilation. In Drew’s snow data assimilation paper he carefully characterized uncertainty in both model simulations and observations, and demonstrated how advanced ensemble data assimilation methods could effectively constrain model simulations in the mountain ranges in the western USA. Drew’s data assimilation work formed the basis for related work on assimilating satellite snow-covered area and streamflow observations in distributed hydrologic models. Drew’s work on mountain hydrology continued throughout the years – important recent papers include Drew’s critical evaluation of methods for snow reconstructions, and his comprehensive compilation and analysis of solar radiation data from mesoscale observing networks. To extend his research in mountain hydrology to adventure-seekers, Drew developed and maintained the popular west-wide real-time snow monitoring website to help outdoor enthusiasts select optimal locations for their winter sport activities.
The highlight of Drew’s career is his work to improve model parameterizations of land processes in the Arctic, and to understand the degradation of permafrost in the 20th and 21st centuries. In 2005 Drew published analyses of permafrost simulations from a coupled climate model, arguably the first study from a climate model that explicitly represented dominant frozen soil processes. The results of this study projected substantial degradation of near-surface permafrost, heightening the concern that permafrost degradation could activate a large portion of the massive Arctic soil carbon pool and accelerate climate change. In this and subsequent papers, Drew exposed land model weaknesses, especially the need to increase the depth of the bottom boundary of the land model and explicitly represent soil organic matter. Drew collaborated with the land modeling team at NCAR for several years in order to address these weaknesses, and this work substantially improved the representation of cold region terrestrial processes in the Community Land Model. Drew’s recent paper on permafrost projections in climate models, published in 2013, provides a robust picture of permafrost degradation in the 21st century, with permafrost most probable only in the Canadian Archipelago, Russian Arctic coast, and east Siberian uplands under a “business-as-usual” emissions scenario.
In remembering Drew, we of course respect his scientific papers – yet, more profoundly, we appreciate Drew’s passion for excellence and his good humor. Drew was vociferously critical of both mediocrity and showmanship. He would always call it as he saw it, no matter what the consequences would be for his professional development. Drew’s public critiques (and the fireworks that accompanied them!) were almost always well-founded, well thought through, and helped many in the community lift their game. He was an excellent mentor for younger scientists, taking the time to teach effective research methods and programming practices. For those of us who had the good fortune to work with Drew, we will always remember his incredible sense of humor, his infectious laugh, and his remarkable ability to develop meaningful personal relationships with most of his colleagues. We have all benefited from Drew’s participation in our projects, in part from his insights and in part because he just made the work more fun.
Drew Slater, land modeler extraordinaire: We miss you.