2012 Progress Report: Forecasting and Evaluating Vulnerability of Watersheds to Climate Change, Extreme Events, and Algal BloomsEPA Grant Number: R835203
Title: Forecasting and Evaluating Vulnerability of Watersheds to Climate Change, Extreme Events, and Algal Blooms
Investigators: Stevenson, R. Jan , Hyndman, David , Moore, Nathan , Qi, Jiaguo
Institution: Michigan State University
EPA Project Officer: Hiscock, Michael
Project Period: June 1, 2012 through May 31, 2017
Project Period Covered by this Report: January 11, 2012 through December 31,2013
Project Amount: $749,801
RFA: Extreme Event Impacts on Air Quality and Water Quality with a Changing Global Climate (2011) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Water Quality , Climate Change , Air , Water
The goals of this project are to advance our knowledge of relationships between extreme events and water quality across the diversity of climatic and geologic conditions of the United States, as well as to develop tools for advancing that knowledge and informing water quality management strategies under different climate change scenarios. We hypothesize that effects of extreme events will be greatest in hydrologically variable regions of the country, which are often most dependent on surface waters for drinking water supply.
To determine the effect of extreme rainfall events on water quality, we are focusing on phytoplankton blooms as our element of water quality. We have tested models for using Landsat satellite imagery to predict phytoplankton abundance (measured as chlorophyll a) in lakes, but found that current modeling approaches could not be calibrated with sufficient precision to be valuable in ecological studies. We tried a new approach, boosted regression tree analysis, using all bands and band ratios of Landsat imagery and developed models for chlorophyll a and Secchi depth (a measure of water transparency). Models for these parameters were developed and tested using data from Landsat images and measured water quality data from 447 lakes sampled during the 2007 National Lakes Assessment of the U.S. Environmental Protection Agency. We found that boosted regression tree models performed much better than traditional regression approaches. As a proof of concept that the remote sensing inferred phytoplankton biomass could be used in ecological studies, we compared regression relationships between chlorophyll a and total phosphorus in lakes when relationships used measured chlorophyll from water samples and inferred chlorophyll from Landsat images. The adjust R2 for the relationship between measured chlorophyll and total phosphorus concentrations was 0.58 and for the relationship between satellite-inferred chlorophyll and total phosphorus concentrations was 0.56. We concluded that satellite images could be used to estimate chlorophyll sufficiently well to be used in ecological studies.
We refined nutrient loading models that have been developed by our group, using both process-based hydrologic models and statistical models. These models predict phosphorus and nitrogen loading from rivers into lakes on daily time intervals based on inputs of weather data, land use, geology, soils, and elevation. The predicted nutrient concentrations and loads by models correlated well with measured concentrations and loads over multiple seasons of the year. The final proof that we can link these models of nutrient loading to phytoplankton blooms in coastal zones of large lakes was tested in the Great Lakes, where we found phytoplankton biomass in coastal zones of the Great Lakes was correlated with nutrient loading from the river and distance from the river mouth, and that 1-week lags between nutrient loading and phytoplankton blooms were particularly significant. Therefore, we conclude that we will be able to use these nutrient loading approaches to forecast nutrient loads under future weather and land use conditions, which we expect to fall within the range of conditions tested in the models.
Future activities will refine the nutrient loading and satellite imaging models for 12 lakes in different climatic and geological regions of the United States. Using nutrient loading and historic satellite images from lakes, we will develop models of algal bloom responses to nutrient loading events. Then we will forecast climate and rainfall changes in the future in these regions, which we will use to drive algal blooms models and determine vulnerability of lake water quality to climate change and extreme events in different geoclimatic regimes.
We plan to develop the website with new results in the next 2 months.