Return to Previous Menu
Return to Previous Menu
Image Formats
AVHRR Satellite Products
What Is Water Reflectance?
Limitations Concerning Reflectance Images
What Is Sea Surface Temperature?
Limitations Concerning Sea Surface Temperature
Images
The satellite products are available in two image formats, GIF and GeoTIFF, to facilitate viewing from a web browser (GIF format) and use in a Geographic Information System (GIS) or image processing system (GeoTIFF format).
The GIF images work well in browsers and almost all image viewers. They are intended to provide a quick general view of the data, but are not suitable for data extraction. For enhanced visual appearance, a land mask has been added to the GIF images.
The GeoTIFF images are suitable for data extraction and have georeferencing information to allow a GIS to determine the earth location of each point. Software such as a GIS or image processing system can generally supply a higher resolution land mask than the images could support, therefore a land mask has not been added to the GeoTIFF files. A GeoTIFF version of the land mask is available.
Top of PageTwo satellite ocean products have been generated from the Advanced Very High Resolution Radiometer (AVHRR) for the period from July 1985 to May 1999. The reflectance or turbidity product is derived from the visible channel of the AVHRR and provides a proxy measure of turbidity associated with suspended sediments. It can also be empirically related to the diffuse attenuation coefficient and Secchi disc depth. The sea surface temperature (SST) product is derived from a "split window" technique on the channels 4 and 5 thermal bands. More information can be found in the Imagery Metadata. An overview of how remote sensing works and its applications can be found in the Remote Sensing Overview.
Top of PageWater reflectance is the proportion of light entering the water that is reflected back out of the water. This property depends on the optical characteristics of water absorption and scattering. Scattering is caused mostly by particles in the water (although pure water scatters blue light, hence the blue color). Light absorption is primarily due to water itself (absorbs red light) and phytoplankton pigments (primarily blue light absorption by chlorophyll). Highly reflective waters are those with a large number of scattering particulates in them. In both highly absorbing and highly scattering waters, visibility into the water is reduced. The highly scattering (turbid) water still reflects a lot of light while the highly absorbing water, such as in a black water lake, is very dark. The scattering particles that cause the water to be turbid can be composed of many things, including sediments and phytoplankton. In an area like the Mississippi River plume, river-borne sediments produce most of the reflectance. The water reflectance derived from a satellite can be used to provide estimates of approximate water turbidity and suspended sediment concentration through comparison with field measurements and in-situ sampling.
From a satellite, a proxy measurement of the water turbidity can be made by examining the amount of reflectance in the visible region of the electromagnetic spectrum. For the AVHRR, the logical choice is channel 1, covering wavelengths 580 to 680 nanometers, the orange and red. In order to make derived products that are comparable over time and space, an atmospheric correction is required. To accomplish this, the effects of Rayleigh scattering are calculated based on the satellite viewing angle and the solar zenith angle, and then subtracted from the channel 1 radiance. For an aerosol correction, channel 2 in the near infrared (750-1000 nanometers) is used. It is first corrected for Rayleigh scattering and then subtracted from the Rayleigh corrected channel 1. The Rayleigh corrected channel 2 is assumed to be aerosol radiance because no return signal from water in the near infrared is expected since water is highly absorbing at those wavelengths. Because channels 1 and 2 are relatively close on the electromagnetic spectrum, one can reasonably assume their aerosol radiances are the same. In addition to the atmospheric scattering corrections, the satellite sensed radiance is also corrected for sun elevation, atmospheric transmission losses, and haze. A complete description of the AVHRR water reflectance derivation can be found in Stumpf and Pennock (1989).
In the included processed imagery, the reflected light emerging from the water column ranges from 0 to 12 percent. The reflectance percentage can be correlated to light attenuation, Secchi disc depth, or total suspended solids, although the exact relationship will vary regionally and depends on the optical properties of the water, which vary widely in the northern Gulf of Mexico coastal region. In the Mississippi River plume region one study has shown that 5 percent reflectance corresponds to a sediment concentration of approximately 130 milligrams/liter (Walker, 1996). The relationship is not linear, so that a reflectance of 2.5 percent was found to correspond to 20 milligrams/liter. Different results can be expected for Galveston Bay, Mobile Bay, the Apalachicola River region, and the west Florida shelf since each of these regions has unique optical properties that can also vary temporally. For additional information on Mobile Bay, see Stumpf (1992).
Top of PageOnly a small fraction of the light incident on the ocean will be reflected and received by the satellite. The probability for a photon to reflect and exit the ocean decreases exponentially with the length of its path through the water because the ocean is an absorbing medium. The more ocean a photon must travel through, the greater its chances of being absorbed by something. After absorption, it will eventually become part of the ocean's heat reservoir. The absorption and scattering characteristics of a water body determine the rate of vertical light attenuation and set a limit to the depths contributing to a satellite signal. A reasonable rule of thumb is that 90 percent of the signal coming from the water that is seen by the satellite is from the first attenuation length. For wavelengths in the near infrared and longer, the penetration depth varies from a meter to a few micrometers. For channel 1, the penetration depth will usually be between 1 and 10 meters. Thus, in shallow clear water, light reflection off the ocean bottom (bottom albedo) will contribute to the total water reflectance seen by the satellite.
Most coastal regions in the northern Gulf of Mexico are sufficiently turbid that the reflectance seen by the satellite has little, or no, component due to bottom albedo. Bottom albedo contains no water column information and is considered a contamination of the signal of interest (water reflectance). There are regions, however, where the bottom albedo may be significant, such as in clear water regions of the west Florida shelf. Therefore, care must be taken in interpretation of the reflectance images in some regions.
Clouds are also problematic in the interpretation of satellite-derived reflectance. Cloud removal algorithms perform a satisfactory job for pixels that are fully cloudy. Partially cloudy pixels are much harder to identify and typically result in false high reflectance estimates. Atmospheric haze is also often not flagged by the cloud algorithm and can result in abnormally high reflectance. A good rule of thumb is that high reflectance values near clouds are suspect.
Some of the images show a bias of high reflectance near the edge of the satellite swath. This can give the erroneous impression that substantial areas of turbid water exist in regions where clear water is expected. Since the satellite is looking through a thicker atmosphere at the limb versus the nadir view, it can be expected that part of the bias could be due to noncompensated atmospheric effects such as increased aerosol scattering. Sun glint is another possibility in some cases.
Top of PageA general definition of Sea Surface Temperature (SST) is the water temperature at 1-2 meters below the sea surface. However, there are a variety of techniques for measuring this parameter that can potentially yield different results because different things are actually being measured.
The earliest technique for measuring SST was to dip a thermometer into a bucket of water manually drawn from the sea surface. The first automated technique for determining SST was accomplished by measuring the temperature of water in the intake port of large ships. This measurement is not always consistent, however, as the depth of the water intake as well as exactly where the temperature is taken can vary from vessel to vessel. Probably the most exact and repeatable measurements come from fixed buoys where the depth of water temperature measurement is always approximately 1 meter, and very robust electrical temperature probes are used. These measurements are usually beamed to satellites for automated and immediate data distribution. A large network of coastal buoys in U.S. waters is maintained by the National Data Buoy Center (NDBC).
Since the 1980s satellites have been utilized increasingly to measure SST and provide an enormous leap in our ability to view the spatial and temporal variation in SST. The satellite measurement is made by sensing the ocean radiation in two or more wavelengths in the infrared part of the electromagnetic spectrum, which can then be empirically related to SST. These wavelengths are chosen because they are (1) within the peak of the blackbody radiation expected from the earth and (2) transmit well through the atmosphere. The satellite measured SST provides both a synoptic view of the ocean and a high frequency of repeat views, allowing an examination of basin-wide upper ocean dynamics that is not possible with ships or buoys. For example, a ship traveling at 10 knots would require 10 years to cover the same area a satellite covers in two minutes.
However, there are several difficulties with satellite-based absolute SST measurements. First, because all radiation emanates from the top "skin" of the ocean, approximately the top 0.1 millimeter or less, it may not represent the bulk temperature of the upper meter of ocean. The "skin" temperature may be either warmer or cooler than the bulk temperature due to the effects of solar surface heating, back radiation, sensible heat loss, and surface evaporation. This discrepancy can make it difficult to compare satellite measurements to measurements from buoys or shipboard methods, complicating ground truth efforts. In addition, satellites cannot look through clouds, creating a "fair weather" bias in the long-term trends of SST. Nonetheless, these difficulties are small compared to the benefits in understanding gained from satellite SST estimates.
Top of PageBesides the aforementioned problem of the skin versus bulk temperature, the SST imagery has an additional limitation during the summertime. Between approximately June through September the surface temperature of the entire Gulf of Mexico warms to a uniform level becoming nearly iso-thermal. There is little or no thermal difference between coastal regions and offshore, making it difficult to impossible to discern water mass changes in the SST imagery. During these periods, the reflectance imagery can still be used to provide information on water clarity.
Top of Page Return to Previous
Menu Return to
Main Menu