REMOTE SENSING DATA SOURCES - CIVILIAN AND NSS

35
Figure 3

36
Measurements of phytoplankton biomass are commonly used to assess the trophic status of lakes and monitor responses to nutrients (George 1997). This is often accomplished by taking
in situ water samples and extracting the photosynthetic pigment, chlorophyll a, with laboratory methods (George 1997). However, Chlorophyll a can also be measured by multi-spectral remote sensing techniques and these techniques have been used successfully in a number of applications
(Bukata et al, 1985, Ramsey and Jensen 1990, Ramsey 1992, George 1997).
Multi-spectral remote sensing methods are based on the fact that phytoplanton, containing chlorophyll a, strongly absorbs energy in the blue and red regions of the electromagnetic spectrum, and reflects energy in the green part of the spectrum (Lo, 1986). By using a basic green/blue band ratio technique, many research applications have successfully correlated in situ
measures of phytoplankton biomass with data derived from data acquired multispectral remote sensing systems. Successful applications have used data from the CZCS, Landsat TM and Landsat
MSS. However, these methods rely on simultaneous in situ phytoplankton measures for calibration. In the Priobskoye study area this type of measurement was not performed at the time of the Landsat TM data collection. Therefore, two other techniques were used to assess potential differences in lake productivity in this area.
The first technique used a simple 2/1 band ratio from 1984 and 1996 Landsat Thematic
Mapper scenes that were acquired for this study. Since airborne oil deposition is not likely to travel long distances, it was assumed that oil effects on lakes would be restricted to areas surrounding the specific oil production sites. Since the TM scene covers an extensive area of landscape, and represents before and after periods of oil activity in the area, any adverse affects on water quality are likely to be restricted to areas surrounding the oil production areas. Green-blue band ratios from both the 1984 and 1996 data showed no significant differences in the band ratio signature from any lakes located throughout the TM scenes, except for areas where there was a significant haze problem in the 1984 imagery and one small lake in the southeast part of the scene.
The second method utilized was a Change Vector Analysis (CVA) technique, which is a radiometric change analysis algorithm that uses multiple dates of geometrically registered and radiometrically corrected imagery (Johnson et al 1997). CVA utilizes n-dimensional multispectral imagery analysis to produce two fundamental statistics from the radiometric comparison of the multiple date images; change direction and change magnitude. These two statistics, when mapped on a Cartesian coordinate system, essentially reduce multiple bands and multiple dates of imagery into a two-dimensional 'change space'. This technique has the advantages of including all multispectral bands in the change determination and can detect changes in both the actual land cover as well as in subtle changes in condition.

37
This CVA technique was applied to the TM data in the overall region of the oil and gas study. Again, very few significant changes in the lake reflectance were noted throughout the greater oil and gas production area and the overall TM scene in general. One small lake in the extreme eastern section of the study area (see Figure 4) appears to have been impacted from sedimentation, most likely from adjacent pipeline and/or tank construction activities. This conclusion was based on the observation of new construction in the area and the fact that this was the only lake where any significant change could be detected. To better validate this conclusion, a statistical comparison of natural fluctuations (using many remote and undisturbed lakes) would show whether the observed change was statistically significant. This is another area where NSS
data can help validate the civilian change detection data for monitoring pre-existing conditions and regulating compliance.
SPOT
SPOT has two operational capabilities. The panchromatic channel is a single-band 10-meter resolution sensor. It has stereo capability, but its 10-15 meter elevation accuracy is of limited value over the flat flood plain. The multispectral sensor on SPOT is similar to Landsat, but its operational period is more limited for change detection work. The 1995 panchromatic image of the study area
(Figure 3) has sufficient resolution to detect pads and pipelines in the developed left-bank region.
Until the new generation of high resolution civilian sensors are launched SPOT panchromatic images are the highest resolution civilian satellite images routinely available.
AVHRR
The Advanced Very High Resolution Radiometer (AVHRR) has been a constant component of the U.S. NOAA weather satellites. The coverage is daily and the resolution is 1.1 km. The Ob
River floodplain is wide enough that it is resolved on the low-resolution AVHRR images. This sensor is a capable of monitoring the ice-blocked northern region of the Ob which causes the extensive flooding at the Priobskoye location. Examples of 1996 AVHRR data from the study area are shown in Figure 5. The April 21, 1996 image shows the snow-covered flood plain which stretches from the upper left to the lower right in the image. By the first of June, 1996, the snow has melted and the false colors from AVHRR bands 2, 5, and 7 show the vegetation (red) and flood conditions (dark blue).

38
Figure 4

39
Figure 5

40
Other Civilian Sensors (SSM/I, ERS-1, JERS-1, Radarsat)
The Special Sensor Microwave Imager (SSM/I) senses microwave radiation emitted from the earth’s surface (i.e., brightness temperatures) at four frequencies, 19.3, 22.2, 37.0, and 85.5
GHz, with vertical polarization at 22.2 GHz and vertical and horizontal polarizations at the other three frequencies. The spatial resolution of the SSM/I is approximately 25 km at 19.3, 22.2, and
37.0 GHz and 12.5 km at 85.0 GHz. The active portion of the SSM/I viewing area covers a swath of 1500 km. SSM/I allows snow coverage and depth to be monitored on the spatial scale of the entire Ob River basin. Snow depth could potentially be monitored using algorithms calibrated with in situ data.
The active radar sensors, ERS-1, JERS-1, and Radarsat, are all-weather sensors that can detect spills during the flood season because of the oil-caused damping of wind-generated waves which changes the radar reflectivity. Active microwave sensors are also used extensively for ice monitoring. ERS-1 data has been used to track the ice floes in Ob Bay. This is important because the time of ice break-up at the mouth of the Ob River determines the flood release in the middle course of the Ob. Flooding has also been tracked on the Ob using ERS-1 data.
National Security Systems
As agreed to by Vice President Gore and Premier Chernomyrdin, the purpose of the EWG
is to examine using any type of national security data acquisition system—space-based, airborne,
oceanographic, or in situ—and derive unclassified products from its data. Because of the remote,
inland location of the Priobskoye site, imaging sensors (both space-based and airborne) fulfilled the above directive for this project.
Other Data Sources—‘In Situ’ and Laboratory Studies
In addition to remotely sensed data, risk assessments require ‘in-situ’ and laboratory data in order to adequately describe the stressor, receptor, and the environment (natural and man-made).
In particular, the risk calculation requires the probability of the stressor occurrence (such as a spill),
the fate of the stressor (e.g. petroleum products) in the environment, and the stressor’s effect on the receptor. The probability of spill occurence requires engineering construction data and failure rate data. Although we have found that the Internet greatly facilitates collaborative projects by making time sensitive data available and communications rapid, we have also found that some data (e.g.
pipeline leak probability) does not exist on the Internet. For instance, it is not available from the
American Petroleum Institute website. There is some useful ‘in situ’ environmental data available on the Internet which is helpful for determining the fate of the stressor. This includes meteorological and river discharge data from the World Meteorological Organization. Even after the stressor fate and

41
transport is determined, its effect on the valued organisms or ecostructures are generally only available from laboratory studies although field data from similar stessor occurances is useful validation data. A useful source for stressor effects on fish is the EPA’s toxicology tabulated data available from the EPA Duluth WEB site.

Download 361.05 Kb.

Share with your friends: