Skip to main content

Framework for selecting climate models for impacts studies in the Northeast

Wednesday, November 16, 2016 | 10:30 am
134 Morrill Science Center
Ambarish Karmalkar
UMass Amherst

Decisions pertaining to the management of natural resources in a changing climate rely heavily on climate change projections produced by dynamical models. While it is difficult to pick ‘best’ climate models from the climate modeling perspective, the impacts community is typically interested in using a handful of models in impacts research due to enormous computational and data storage costs. In this talk, I will describe a framework - a two-step process - developed to select a subset of credible climate models (global and downscaled) that can be used for impacts studies in the Northeast. The first step is to evaluate the performance of these models in simulating historical climate of the Northeast. In addition to evaluating models’ ability to capture seasonal temperature and precipitation climatologies, we also evaluate how successful these models are in simulating large-scale circulation features that drive the climate of the Northeast. Such a process-based assessment is crucial to establish credibility of models' future projections. The other step is to make sure that the selected models provide a diverse set of climate futures consistent with uncertainty in model projections. This approach will allow us to characterize and categorize model projections for the Northeast and will take us a few steps closer towards answering the question: which climate models should I use in my work? The model selection framework is developed in collaboration with Ray Bradley (UMass Amherst), Alex Bryan (USGS), and Jeanne Thibeault and Anji Seth (UConn).


Ambarish Karmalkar is a research fellow at the NE CSC at UMass Amherst. Ambarish’s research focusses on producing and communicating physically consistent and spatially coherent information on climate change and associated uncertainties for impacts assessment. For his Doctorate in Geosciences at UMass Amherst, Ambarish used a regional climate model to study climatic changes in Central America. Ambarish did his postdoctoral work at Oxford University, UK, where he studied climate change projections for the island nations in the Caribbean. He then worked as a research scientist for thee years at the Met Office Hadley Centre, UK, focusing on developing methods to quantify uncertainties in climate model projections. As an undergraduate Ambarish studied Physics at the University of Mumbai, India and received a Masters in Astronomy from UMass Amherst.