"Proceedings, Federal Interagency Workshop,
"Sediment Technology for the 21'st Century,"
St. Petersburg, FL, February 17-19, 1998"

Mapping Suspended Sediments Using Remotely Sensed Satellite Images:
San Francisco Bay

By Pat S. Chavez, Jr.

Abstract

The large areal coverage of remotely sensed satellite images makes them useful for mapping and monitoring large regions and studying entire ecosystems. This can be important in helping understand how various parts of an ecosystem react to environmental stress introduced at different locations and/or to climatic differences occurring both temporally and spatially. Along with the large regional coverage of satellite image data the repetition aspect of these data makes them applicable to monitoring a variety of environments on the Earth's surface, including coastal waters. Satellite images of coastal waters offer the potential to detect and map the amount of suspended sediment residing in a given system, and can help with the analysis of how much sediment is entering and leaving a given area.

For this study of San Francisco Bay, we used Landsat Thematic Mapper (TM) images (30-meter horizontal resolution) and field spectral radiometer data collected during several water-sampling cruises conducted by the USGS Water Resources Division (WRD). The field spectral radiometer has wavelengths identical to those on Landsat TM and its data were used with the water sample results to build relationships between spectral reflectance and both suspended particle matter (SPM) and chlorophyll-a concentrations. The near-infrared and red spectral bands were better than the green and blue bands for mapping medium to relatively high SPM concentrations, with the near-infrared band being the best for overall correlation with both SPM and chlorophyll-a. The derived relationships of both SPM and chlorophyll-a versus the radiometer near-infrared spectral reflectance were used on the near- infrared Landsat TM image that had been converted to spectral reflectance. An advantage of using the near-infrared spectral band is that because of its minimal water penetration most sub-bottom reflectance problems, especially in clear and shallow waters, are eliminated. This is not the case when using the blue or green spectral bands. Because of the minimal water penetration of the near-infrared band computed SPM and chlorophyll-a concentrations represent only the surface waters or upper one meter. To get a more representative sampling of the water column the blue and green bands could be used because of their greater water penetration capability, but in this study both bands had problems mapping medium to high levels of SPM concentrations.

Multi-temporal Landsat TM images were used to compute the number of tons of SPM and chlorophyll-a over the entire bay for the upper most one meter of water. Then using WRD's water samples collected at one-meter intervals from the surface to the bay floor we analyzed the vertical variability and computed a multiplicative coefficient for use with the Landsat TM derived digital SPM and chlorophyll-a image maps. This allowed us to compute the total SPM and chlorophyll-a volumes for the entire bay.

In general, the reflectance of water increases with increased suspended sediment concentrations (positive correlation) and decreases with increased salinity (negative correlation). Mapping suspended sediment concentrations and chlorophyll-a using satellite images has been described in previous studies. However, most previous studies have not attempted to tackle the temporal comparison and change detection of the water parameters, mainly because of the difficulty associated with absolute radiometric calibration of the satellite images. From a monitoring point of view change detection, therefore calibration, is a critical aspect of ecosystem evaluation and management.

A major goal of this project has been to use spectral reflectance measurements made in the field during water sampling cruises, along with a satellite calibration and radiometric correction model to convert satellite digital numbers (DNs) to surface reflectances, allowing the mapping of desired water parameters on a temporal basis without the need for water sampling during each satellite overflight. Our satellite radiometric correction model makes corrections for sensor gains and offsets, spectral irradiance, solar elevation, atmospheric scattering and absorption (additive and multiplicative effects), and Earth-Sun distance. The SPM and chlorophyll-a values computed from the water samples and the spectral reflectance measurements made in the field during the water sampling were used as input to regression analysis. The resulting relationship was used to transform the satellite surface reflectance images, which are generated using our radiometric calibration and correction model, into SPM and chlorophyll-a digital image maps. We have generated digital SPM image maps and corresponding statistical information using several multi-temporal Landsat TM images. The results are being used to study and compare suspended sediment volumes of the entire Bay, as well as compare various regions of the Bay both spatially and temporally. The computed amount of Bay wide SPM volumes for the given image dates range from 230,000 to 450,000 tons, while the surface concentration levels vary from less than 5 mg/l to over 250 mg/l. At this time a Landsat TM image collected during the winter high runoff season has not been acquired and analyzed, so the range in both total SPM volumes and concentrations is expected to be larger when one is included. The digital SPM image maps visually show the surface suspended sediment patterns and horizontal spatial variability within the Bay.


upList of Participants


[an error occurred while processing this directive]
[an error occurred while processing this directive]