Each year, AGCI hosts several public talks featuring leading global change researchers and practitioners. Presented by participants in AGCI’s signature workshop series, these lectures cover the gamut of global change topics from biodiversity threats to urban heat resilience to the history and future of Earth’s climate trajectory. AGCI’s public lecture series honors Walter Orr Roberts (1915-1990), noted humanitarian, scientist, and founder of the National Center for Atmospheric Research (NCAR).
To improve societal and environmental wellbeing, decision makers need climate information that is actionable. Actionable climate information needs to be timely; sufficiently detailed as to describe the region and processes of interest; trustworthy and truthful in acknowledging uncertainties; and policy- or decision-relevant. We have already begun to observe the transformative power of machine learning (ML) to produce actionable knowledge in weather forecasting (where we can conveniently check whether our predictions are correct). How could machine learning do the same, and more, in the context of climate change? This talk will introduce how we make climate predictions, and then describe ways in which machine learning can enable us to make more trustworthy and more detailed climate predictions, at a faster pace. We'll also explore how machine learning can provide a bridge between climate.
The Colorado River is critically important--it supplies water to over 40 million people, irrigates over 5 million acres of agriculture, and supports hydropower, environmental, and recreational resources. The Colorado River Basin is also incredibly complex--it spans seven U.S. states and two Mexican states, has highly variable hydrology, and is overallocated. Long-term planning in the Colorado River Basin has always been challenging due to uncertainties in hydrology, demand, policy, and different management priorities among stakeholders. These challenges are now exacerbated by the need to account for potential impacts of climate change. This context is best described as deep uncertainty, where a wide range of assumptions about future conditions are plausible, multiple management perspectives are expressed, and it is impossible to identify the best assumptions about conditions or priorities.
This talk presents studies conducted by Bureau of Reclamation's Colorado River Basin Modeling and Research Team that demonstrate uncertainty in climate and hydrology and explore Decision Making under Deep Uncertainty (DMDU) techniques to help address planning challenges.
Urban heat is deadlier than nearly all other U.S. weather-related hazards combined, with risks increasing due to climate change and the urban heat island effect. During this summer's record-breaking temperatures, urban heat is at the forefront of the national conversation on climate risk, intersecting with and compounding the COVID-19 pandemic, social inequity, and racial injustice. Fundamental research continues to advance understanding of the characteristics of resilient cities and their governance, but translating this knowledge about urban resilience into practice remains a challenge.