Communities throughout New England are being impacted by rapid development and associated land use changes while facing serious impacts from a changing climate. Together, these affect frequency and magnitude of flood events, droughts, nutrient loading, and other characteristics that determine water quality and quantity in the region. Individuals, all levels of government, not-for-profits, and the private sector are seeking decision relevant information to help them adapt to existing and future risks and uncertainties that accompany ongoing environmental change. Chief among the information needed for adaptation are assessments of possible ecological, social, and economic consequences to critical watershed services.

Here we discuss the development of climate assessment reports for two coastal watersheds in New England: Casco Bay and Great Bay.

The Great Bay National Estuarine Research Reserve (GBNERR) is one of twenty-eight research reserves in a national network of protected areas established for long-term research, education and stewardship. The Great Bay estuary is New Hampshire's (NH) largest estuarine system and includes 30 km² of open water, wetlands, and diverse habitats including eelgrass beds, mudflats, saltmarsh, rocky intertidal, and upland forests and fields.

The Casco Bay Estuary in southern Maine (ME) is an Environmental Protection Agency designated 'estuary of national significance'. It encompasses over 2500 km², 25% of the state's population, and the state's largest metropolitan area, Portland. In addition, the watershed boasts 785 islands, islets, and exposed ledges, and more than 2180 km of rivers and streams.

We evaluate historical climate indicators over the past century including annual and seasonal temperature and precipitation, temperature and precipitation extremes, winter snowfall and snow cover, peak river flow volume and center of volume date, lake ice-out dates, and sea surface temperature (SST) of the Gulf of Maine. Temperature, precipitation, snowfall, and snow depth data are provided by the United States Historical Climate Network V2 dataset and from the US Cooperative Summary of the Day (DSI-3200). River flow data and lake ice-out data come from the United States Geological Survey. Sea surface temperature data is the average of three 2°x2° Extended Reconstructed SST grid cells covering the Gulf of Maine. Historical trends are calculated using Sen's slope and statistical significance using the Mann-Kendall nonparametric statistic.

Both watersheds exhibit statistically significant warming trends in annual maximum temperature ranging from +0.06°C/decade to +0.12°C/decade over the period 1895-2009. Over the period 1970-2009, winter minimum temperatures exhibit the fastest seasonal rate of warming, ranging from +0.49°C/decade to +0.63°C/decade. Since 1887, the date of lake ice-out on Lake Winnipesaukee in NH has become 5 days earlier, and almost 20 days earlier on Sebago Lake, ME. No statistically significant trends in snow cover or snowfall were detected over the period 1949-2009. Over the period 1895-2009, statistically significant increasing trends in annual precipitation are driven primarily by statistically significant increases in fall precipitation. Increases in fall precipitation may be strong contributors to observed statistically significant increases in fall discharge on the Lamprey and Oyster Rivers. Trends in Gulf of Maine SSTs are significant at the annual and seasonal level for all seasons. Over the period 1887-2008, spring SSTs have risen the fastest at 0.13°C/decade.

A modified statistical asynchronous regression (SAR) downscaling approach uses daily predictor fields from four coupled atmosphere ocean general circulation models (AOGCMs) to statistically downscale maximum and minimum temperature and 24h cumulative precipitation through 2099 under higher (A1Fi) and lower (B1) emissions scenarios. Model predictor values and observed predictand values are ranked and a piecewise linear regression function is fitted to the datasets by month, including two weeks of overlapping data on either side. The most reliable predictor for daily maximum and minimum 2m temperature were those same fields as simulated by the AOGCMs. The downscaling model for precipitation is similar to that for temperature in many aspects, but with some key differences. First, for practical reasons an AOGCM predictor had to be chosen that was commonly archived at the daily scale. Although upper-level humidity and geopotential height have shown promise in downscaling precipitation, few AOGCMs have preserved daily outputs. Thus, 24-hr cumulative precipitation was selected as the predictor for precipitation, with the additional refinement of incorporating convective and large-scale precipitation if both predictors were available. For models with these variables, the downscaling approach selects from three possible predictors the one best suited to each month: convective, large-scale, or total. Second, EOF filtering of the GCM output is not performed since we found that to degrade the results along with introducing negative values for precipitation. Finally, the logarithm of precipitation values is used instead of raw precipitation amount to decrease the residuals of the regression.

Temperature increases under the higher emissions scenario are nearly double that expected under the lower emissions scenario by the end of the 21st century (2070-2099). Overall, Great Bay and Casco Bay can expect to see increases in annual average maximum and minimum temperature ranging from +1.5°C to +5.0°C by the end of the century, with the greatest increases occurring during the summer months. The average number of days above 35°C is projected to increase five-fold under the higher emissions scenario during the latter part of the 21st century (2070-2099). Annual precipitation is projected to increase by about 5%-10% under both higher and lower emissions scenarios, albeit with a large range of uncertainty. The frequency of extreme precipitation events is also expected to increase, with larger increases projected under the higher emissions scenario. By late-century, the number of snow-covered days is projected to decrease by 60% under higher emissions, leaving Great Bay with only one month of snow cover in winter.

The results of this study clearly demonstrate that historical and future climate change poses a risk to the GBNERR. One of the most pressing issues impacting the Great Bay is nitrogen loading through non-point sources, which has been shown to increase during periods of heavy rainfall. The observed and projected increases in precipitation documented in this study threaten to exacerbate existing nitrogen pollution in the future. The effects of extreme heat and decreased snow cover on GBNERR habitats need further study.