We analyzed observational temperature and precipitation data from the NOAA Global Historical Climate Network–Monthly gridded products, which are available on a 5° by 5° geographical grid (29). We restricted our observational analysis to grid points for which observational data were available for the full 1931–2015 period, with the joint probability analysis restricted to those grid points for which both observational temperature and precipitation data were available. The NOAA temperature and precipitation data were available as monthly anomalies, which we aggregated to annual values. Because the temperature and precipitation data used different baseline periods for calculating the anomalies (1981–2010 for temperature and 1961–1990 for precipitation), we used a simple arithmetic adjustment to express each of the temperature values as an anomaly from the 1961–1990 period (i.e., by adding the difference between the 1981–2010 calendar-month mean and the 1961–1990 calendar-month mean to each monthly temperature value).

We also analyzed climate model data from the CMIP5 (30). To match the NOAA observational data, we calculated monthly temperature and precipitation anomalies using the 1961–1990 baseline. Following many previous studies, we accessed all available CMIP5 realizations and interpolated all climate model realizations to a common 1° by 1° geographical grid. For the regional analyses (see below), we used all the 1° by 1° grid points in each region.

We used four of the CMIP5 forcing experiments (see tables S1 to S3 for a list of realizations). To analyze historical changes in the probability of warm and dry years, we used the Historical and Natural forcing experiments, which were available through the year 2005. The historical forcing experiment includes both natural and anthropogenic climate forcings during the historical period, while the natural forcing experiment includes only the natural climate forcings. We tested the statistical significance of the differences between the Historical and Natural sample populations using the Kolmogorov-Smirnov test (see the Supplementary Materials). For this comparison, we evaluated two categories of statistical significance: P value less than 0.01 and P value between 0.01 and 0.05.

We also compared the probability of warm and dry years in the 2020–2050 period of the RCP2.6 and RCP8.5 future climate forcing scenarios. Of the RCP scenarios archived in CMIP5, RCP8.5 most closely tracks the recent emissions trajectory, while RCP2.6 most closely represents the ambitious mitigation detailed in the UN Paris Agreement (31) (see the Supplementary Materials for additional details).

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.