EM 1110-2-2907
1 October 2003
and satellite data. The potential use of ALI and hyperspectral Hyperion for studies of forests
in remote locations and forests at risk may greatly enhance forest management and lower the
costs associated with ecological monitoring. Accurate estimates of LAI based on satellite
imagery have the potential to support forest biomass monitoring, and hence forest health and
changes in canopy structure attributable to pollution and climate change. The ability to esti-
mate LAI with remote sensing techniques is, therefore, a valuable tool in modeling the eco-
logical processes occurring within a forest and in predicting ecosystem responses.
Point of Contact: Jerry Ballard, Phone: (601) 634-2946
6-7 Case Study 5: Blended Spectral Classification Techniques for Mapping
Water Surface Transparency and Chlorophyll Concentration
Subject Area: Water quality
Purpose: To establish water clarity and algal growth in a dam reservoir
a. Introduction.
(1) An accurate portrayal of water clarity and algal growth in dynamic water bodies
can be difficult owing to the heterogeneity of water characteristics. Heterogeneity can stem
from the spatial distribution of sediments delivered to a lake by a tributary. Water turbidity
associated with tributary sediment load controls water clarity and subsequently will impact
algae growth. Additionally, algal growth will influence water clarity by reducing water
transparency during times of algal blooms. Both algal growth and sediment turbidity are
controlled by such factors as water depth, flow rate, and season.
(2) To better monitor the water quality at dam reservoirs, a spatial estimate of both
water clarity and algal chlorophyll over a broad area is required. To accurately capture these
properties a large number of water samples must be taken, a task that may not be feasible for
most studies. Remote sensing lends itself well to the assessment of water quality testing at a
variety of spectral scales due to the response of suspended sediment in the visible and ther-
mal spectrum. Chlorophyll, produced by algae, can also be detected by its visible and infra-
red emission. The study reviewed here developed a classification algorithm to predict water
clarity and chlorophyll concentrations. The algorithm was based on a correlation between
spectral data and the empirical field data. Previous studies attempting to classify water clar-
ity and chlorophyll required field sampled training sites. The goal of this study was to de-
velop an algorithm based on empirical data that would illuminate the need for such test
training sites. Thus, researchers testing for water quality would then need only the Landsat
TM data to monitor water quality at a fresh water lake.
b. Field Work. Secchi Disk measurements and water samples were collected at a dam
reservoir in conjunction with a Landsat fly-over at West-Point Lake, Georgia. Water sam-
ples were frozen and stored in a dark room to preserve the algae populations. These samples
where later analyzed for chlorophyll (Ca) concentrations. Water clarity was measured in situ
with a Secchi Disk (Sd). This 20-cm disk estimates water clarity by measuring the depth to
6-12
Previous Page
