Goal: The overall goal is to analyze and interpret all available operational remotely sensed imagery of southern Lake Michigan using state-of-the art image processing techniques.

Objectives:

1. analyze and archive historical Advanced Very High Resolution Radiometer (AVHRR) lake surface temperature (LST) and remote sensing reflectance (R_rs) imagery.
2. provide the first-ever chlorophyll (C_SAT) and R_rs (related to turbidity) estimates from the Sea-viewing Wide-Field-of-View Sensor (SeaWiFS) imagery using a revised atmospheric correction procedure de-veloped by Stumpf et al.
3. create year-long, uninterrupted time-series SeaWiFS imagery using a statistically-based time-space in-terpolation procedure.
4. create multi-sensor visualizations of physical (AVHRR LST) and particle (SeaWiFS C_SAT, R_rs) data.
5. validate the satellite-derived water parameters with in situ data from intensive field surveys at sampling locations.
6. support modeling groups' efforts (Schwab et al., Chen et al.) by providing synoptic satellite water pa-rameters for model experiments.

Approach:

• All available AVHRR (1989 to present) and SeaWiFS (1997 to present) imagery were processed and are being analyzed. Geotiff images are posted at www.geo.mtu.edu/great_lakes/lakersi/cgi-bin/seawifs.cgi and www.geo.mtu.edu/great_lakes/lakersi/cgi-bin/avhrr.cgi.
• Multi-sensor visualization techniques were employed in comparative studies of the time, duration and spatial extent of episodic events using statistically-based interpolation techniques, an edge detection al-gorithm (Diehl and Budd, in review) and data fusion of simultaneously-acquired remotely sensed data.
• A digital bathymetry model was coupled with SeaWiFS sediment maps (as R_rs) to obtain instantaneous estimates of total mass of resuspended sediments in southern L. Michigan.
• Comparisons of ship-based chl-a and TSS with satellite-derived parameters provide validation of Sea-WiFS C_SAT and R_rs.

Selected Results:

• Historical AVHRR Rrs imagery from 1992 to 1999 reveal that episodic resuspension events occur throughout the fall, winter and spring months from from September to the end of April.
• Distinct horizontal temperature gradients from nearshore to offshore were observed in late March and April in Lake Michigan. The presense of a ther-mal bar indicated a transition zone between near-shore (warmer) from offshore (colder) waters.
• Using SeaWiFS chlorophyll and sediment maps of southern Lake Michigan in late spring, we identi-fied two chlorophyll maxima in nearshore and off-shore waters, which were separated by the thermal bar. Whereas the nearshore chlorophyll maxima was strongly coupled to resuspended sediment, the offshore chlorophyll maxima was not.
• Calculations of daily total mass of sediment during the spring plume event of 1998 range from 2.8x10 9 kg (Fig. 1).

Fig. 1. Total mass of sediment/day in southern L. Michigan during spring 1998 based on SeaWiFS interpolated TSS estimates and 1 km 2 bathymetry contours. Gaps in data represent periods when clouds precluded interpolation.