Quantifying uncertainty cascading from climate, watershed, and lake models in harmful algal bloom predictions.

In response to increased harmful algal blooms (HABs), hypoxia, and nearshore algae growth in Lake Erie, the United States and Canada agreed to phosphorus load reduction targets. While the load targets were guided by an ensemble of models, none of them considered the effects of climate change. Some watershed models developed to guide load reduction strategies have simulated climate effects, but without extending the resulting loads or their uncertainties to HAB projections. In this study, we integrated an ensemble of four climate models, three watershed models, and four HAB models. Nutrient loads and HAB predictions were generated for historical (1985-1999), current (2002-2017), and mid-21st-century (2051-2065) periods. For the current and historical periods, modeled loads and HABs are comparable to observations but exhibit less interannual variability. Our results show that climate impacts on watershed processes are likely to lead to reductions in future loading, assuming land use and watershed management practices are unchanged. This reduction in load should help reduce the magnitude of future HABs, although increases in lake temperature could mitigate that decrease. Using Monte-Carlo analysis to attribute sources of uncertainty from this cascade of models, we show that the uncertainty associated with each model is significant, and that improvements in all three are needed to build confidence in future projections.

Bias correction of climate model outputs influences watershed model nutrient load predictions.

Waterbodies around the world experience problems associated with elevated phosphorus (P) and nitrogen (N) loads. While vital for ecosystem functioning, when present in excess amounts these nutrients can impair water quality and create symptoms of eutrophication, including harmful algal blooms. Under a changing climate, nutrient loads are likely to change. While climate models can serve as inputs to watershed models, the climate models often do not adequately represent the distribution of observed data, generating uncertainties that can be addressed to some degree with bias correction. However, the impacts of bias correction on nutrient models are not well understood. This study compares 4 univariate and 3 multivariate bias correction methods, which correct precipitation and temperature variables from 4 climate models in the historical (1980-1999) and mid-century future (2046-2065) time periods. These variables served as inputs to a calibrated Soil and Water Assessment Tool (SWAT) model of Lake Erie's Maumee River watershed. We compared the performance of SWAT outputs driven with climate model outputs that were bias-corrected (BC) and not bias-corrected (no-BC) for dissolved reactive P, total P, and total N. Results based on graphical comparisons and goodness of fit metrics showed that the choice of BC method impacts both the direction of change and magnitude of nutrient loads and hydrological processes. While the Delta method performed best, it should be used with caution since it considers historical variable relationships as the basis for predictions, which may not hold true under future climate. Quantile Delta Mapping (QDM) and Multivariate Bias Correction N-dimensional probability density function transform (MBCn) BC methods also performed well and work well for non-stationary climate scenarios. Furthermore, results suggest that February-July cumulative load in the Maumee basin is likely to decrease in the mid-century as runoff and snowfall decrease, and evapotranspiration increases with warming temperatures.

Climate Change and Nutrient Loading in the Western Lake Erie Basin: Warming Can Counteract a Wetter Future.

In the past 20 years, Lake Erie has experienced a resurgence of harmful algal blooms and hypoxia driven by increased nutrient loading from its agriculturally dominated watersheds. The increase in phosphorus loading, specifically the dissolved reactive portion, has been attributed to a combination of changing climate and agricultural management. While many management practices and strategies have been identified to reduce phosphorus loads, the impacts of future climate remain uncertain. This is particularly the case for the Great Lakes region because many global climate models do not accurately represent the land-lake interactions that govern regional climate. For this study, we used midcentury (2046-2065) climate projections from one global model and four regional dynamically downscaled models as drivers for the Soil and Water Assessment Tool configured for the Maumee River watershed, the source of almost 50% of Lake Erie's Western Basin phosphorus load. Our findings suggest that future warming may lead to less nutrient runoff due to increased evapotranspiration and decreased snowfall, despite projected moderate increases in intensity and overall amount of precipitation. Results highlight the benefits of considering multiple environmental drivers in determining the fate of nutrients in the environment and demonstrate a need to improve approaches for climate change assessment using watershed models.