trenches may also present unique safety risks to the worker and 
environment. While use of imagery will not totally eliminate 
the need for ground sampling, it can substantially reduce the 
amount that is required. Through proper analysis of imagery 
data, it should be possible to locate the buried material with 
greater precision, reduce the ground sampling requirements, 
and ensure greater safety in the clean-up process. 
3. CASE STUDY: CLINCH RIVER 
ENVIRONMENTAL MONITORING PROGRAM 
3.1 Introduction 
The Clinch River is the main receiving stream for point and 
nonpoint sources discharges from the DOE Oak Ridge 
Reservation (ORR). Two major surface water tributaries to the 
Clinch River provide the majority of the contaminant flux from 
DOE sites: White Oak Creek and Poplar Creek. 
Quantifying the impacts of these inflows to the off-site 
environment is a major DOE concern. Specifically, knowledge 
of the spatial extent and hydrodynamics of the inflow mixing 
zones is necessary to adequately design water sampling 
programs for detecting off-site contamination flow by surface 
water and for ensuring that adverse health risks are not present. 
Since the Clinch River is the major integrator of all 
groundwater and surface water contamination from the ORR, 
delineation of inflow mixing zones (spatial extent and temporal 
variations) is required to develop efficient sampling plans to 
monitor actual contaminant levels both prior to remediation of 
onsite waste areas and for long-term monitoring. At the mixing 
zone, contaminant inflows are of the highest concentration (i.e., 
least dilution) and thus may present the greatest risk concern. 
This case study utilized the analysis of remotely sensed thermal 
and visible imagery to assess drainage systems on the DOE 
ORR into the Clinch River. In addition, this study was also 
designed to incorporate both image-derived and in-situ “ground 
truth” information for use in modeling the surface-water 
transport of contaminants. The modeling work is crucial to 
understanding the mixing zones. 
This project proved that, through the use of remotely sensed 
imagery, it is possible to map aqueous mixing processes. 
3.2 Data Collection 
Over the past several years, remote sensing imagery has been 
collected of the DOE ORR by several groups working on 
various environmental problems. This project used remote 
sensing datasets collected by the DOE Oak Ridge Operations 
Remote Sensing Program in 1992 and 1994 and topographic 
datasets collected by the Lockheed Martin Energy Systems 
Geographic Information Systems and Spatial Technologies 
(GISST) Program in 1994 and 1995. 
Since the main goal of this project was to perform a preliminary 
analysis of thermal mixing of the main tributaries to the Clinch 
River from the ORR using remote sensing imagery, it was 
necessary to extract various remote sensing information 
covering the confluences of the tributaries with the Clinch 
River. Some additional watershed analysis was performed 
using digital terrain data. The dataset of most utility was the 
long wave-band thermal infrared imagery, available from both 
night and daytime aerial surveys during April 1992 and March 
1994. These surveys were conducted by EG&G Energy 
Measurements using DOE-owned equipment that included a 
Daedalus 1268 multispectral scanner. During these surveys, the 
confluences of White Oak and Polar Creeks were remotely 
sensed at a spatial resolution of 1.5 to three meters per pixel. 
3.3 Results 
Daedalus imagery collected in 1992 and 1994 was used to 
analyze and delineate the mixing zones at both White Oak 
Creek and Poplar Creek. Figure 2 illustrates the mixing zone 
of the White Oak Creek inflow to the Clinch River, as seen on 
1994 Daedalus thermal imagery. Statistical analysis of the 
imagery for White Oak Creek was also performed to assess the 
distribution of thermal differences in the area of the inflow. 
Analysis of the pixels throughout the mixing zone (starting at 
the source of the inflow and extending 200 meters downstream) 
revealed the results as shown in Table 3. 
The Daedalus imagery indicated that the thermal mixing 
patterns of the tributaries to the Clinch River varied markedly 
from date to date and from night to day. Thermal plumes were 
very prominent in some imagery, allowing ready 
characterization of surface thermal mixing zones. On other 
occasions, surface thermal mixing zones were poorly delineated 
or below the detection limits of the sensor. The dramatic 
differences in thermal mixing patterns from dataset to dataset 
may be expected to be attributable to a number of factors, 
including: 
° day/night differences in thermal inertia; 
. flow and water depth differences of the Clinch River due 
to changes in releases from an upstream hydroelectric dam 
(Melton Hill Dam) 
° differences in velocity and sediment loads of the two 
streams and their tributaries; 
differences in current and preceding meteorological 
events, including precipitation, air temperature, and 
atmospheric parameters; and 
° the three-dimensional character of each stream (water 
depth profiles) in the vicinity of the stream confluences. 
Although it is clear that mixing patterns vary greatly due to a 
combination of factors such as those listed above, these factors 
were not fully evaluated in the initial study and more work is 
required to understand their effect. Thermal imagery can, at 
most, capture the mixing regimes upon a limited number of 
specific occasions. To create an effective water sampling plan 
for monitoring contaminant transport, use of models is essential 
(1) to characterize mixing zones at other times and stream 
conditions, and (2) in order to select both optimal times and 
optimal locations for collection of monitoring samples. 
3.4 Summary 
This project demonstrated an approach that is applicable to 
monitoring any run-off or effluent entering a neighboring body 
of water, provided that the plume exhibits a thermal or spectral 
signature observable on imagery. Some examples of where 
these methods could be used include monitoring of: 
° waste water or cooling water inflows from major 
industrial facilities or power plants; 
° potentially polluted streams or rivers entering a bay, inlet, 
or other coastal waters; 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996