Variability in NEX-GDDP CMIP6 Ensemble for Alaska, American Samoa, Guam, Northern Mariana Islands, and Hawaii

The NEX-GDDP (NASA Earth Exchange - Global Daily Downscaled Projections) dataset is an ensemble of 35 global climate models (GCMs) from the Coupled Model Intercomparison Project, Phase 6 (CMIP6). The GCMs were statistically downscaled using the bias-corrected spatial disaggregation (BCSD) method and trained against weather observations from the Global Meteorological Forcing Dataset (GMFD). The GCMs used in the dataset were produced from a myriad of climate research groups of diverse geographic origin. The dataset consists of the four Tier 1 scenarios from the CMIP6 Shared Socioeconomic Pathways (SSPs): SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5. The NEX-GDDP dataset has 0.25-degree spatial resolution and a daily temporal resolution.

This project focuses on the Pacific US states and territories outside of the contiguous US (OCONUS) to provide information for geographic areas often underrepresented in regional climate datasets. Our climate comparison covers four Pacific regions: Alaska, American Samoa, Guam and Northern Mariana Islands, and Hawaii. Although the NEX-GDDP projections are relatively coarse for these regions, this is the only dataset that covers the Pacific OCONUS.

Here, we present analysis of the NEX-GDDP ensemble spread across four climatological variables: change in total precipitation, change in total potential evapotranspiration (PET), change in maximum temperature, and change in minimum temperature. The changes were calculated between the historical period (1976-2005) and near-term future (2025-2054), mid-term future (2045-2074), and long-term future (2070-2099). The data processing started with downloading the NEX-GDDP ensemble for daily maximum near-surface air temperature, daily minimum near-surface air temperature, and precipitation. The PET was calculated using the Hamon method based on minimum temperature, maximum temperature, and latitude. The variables were then subset to the geographic areas of interest and processed to derive variables for each season (fall, spring, summer, winter) and year. Not all GCMs downscaled in NEX-GDDP had availability for every variable and scenario combination. The ensemble data were next summarized by USGS 12-digit hydrologic unit (HUC-12) and reported as the median, maximum, and minimum of the ensemble. To reduce the effect of outliers, no weights were applied. These derived datasets are included in the Future Scenarios data collection in the EPA EnviroAtlas tool.

The NEX-GDDP dataset across all study areas shows increases in average minimum and maximum temperature. Alaska shows the greatest median change in both temperature variables of any of the Pacific OCONUS regions. Increases in potential evaporation are also projected in all Pacific OCONUS regions. Projected changes in precipitation have much higher variability across different domains. Alaska is projected to experience substantial increases in precipitation, while the Pacific Islands have a small median change with a wide variability across NEX-GDDP GCMs.

The future climate projections generated in this study can inform climate adaptation and remediation discissions by regional planners and managers in the Pacific OCONUS, where availability and access to climate data remains a priority need.