Fig. 3
From: Learning about climate change uncertainty enables flexible water infrastructure planning
Learning over time using virtual observations. a–d One sample realization of Bayesian learning over time in which precipitation increases strongly. Black dots represent a time series of virtual climate observations. Shaded regions indicate the projected 90% CI, updated with each time period’s virtual observation. Virtual observations of T (a) and P (b) are used to simulate MAR (c), and water shortages assuming 80 MCM dam capacity (d). e, h As in a–d but for an alternative realization of virtual observations, showing modest decrease in P
