OVERVIEW:

A changing climate implies a shift in the magnitude of key environmental properties such as temperature, precipitation, and productivity. It also often causes changes in the phenology - the seasonal timing - of features such as peaks or troughs in these properties. In marine and coastal systems, many key ecological events - migrations, reproduction, and exploitation of blooms - rely on the phenology of the underlying environmental processes. The socioeconomic benefits of these ecosystems derived from natural resource extraction / processing and recreation / tourism are also closely tied to the timing of seasonal events. Changes in timing can often be more important than changes in magnitude to the affected ecological processes.  There is a need to incorporate phenology into climate assessments more broadly. We propose a study that (1) develops methodologies for identifying and quantifying key phenological properties for the marine environment, (2) develops tools for visualizing timing-based events spatially, (3) evaluates these indices against important ecological shifts to determine a subset of phenological indices that are most informative to science and management, and (4) prototypes forecasting programs for the most informative indices.

(1) Methods for extracting phenological properties from seasonal time series: We will develop and test robust analytical methods for identifying and quantifying key phenological properties from satellite data that can be applied generally to non-stationary time series with both simple and complex seasonal patterns.

(2) Calculation/visualization of phenological index time series from satellite-measured fields: We will use the methodologies from part (1) to calculate phenological index time series for key oceanographic variables (sea surface temperature, precipitation, wind speed, chlorophyll-a, sea ice properties) and additional derived indices.

(3) Evaluation of indices by comparison with ecological time series: We will evaluate these indices against ecologically and socioeconomically important time series: return dates of diadromous fish, ramp-up dates of fisheries (e.g. the Maine lobster fishery), long running time series of zooplankton and ichthyoplankton, and occurrence of marine pathogens.

(4) Development of operational forecasts:   In many cases, phenological indices are more amenable to forecasts than state variables. We will develop forecasting algorithms for those indices that emerge as most informative (from part 3), with the objective of providing useful information to stakeholders in these ecosystems.