261 
Densitométrie Analysis of 37 Lakes Using 70mm ERTS Imagery 
The primary goal of this study is to evaluate the usefulness of 
densitométrie analysis of ERTS photographic imagery as a tool for a 
periodic monitoring program of Wisconsin lakes for changes in water 
turbidity caused by the growth of phytoplanktonic algae. Thirty-seven 
lakes were selected for densitométrie analysis in each of the four ERTS 
bands. The location of these lakes, which range in fertility from 
extremely eutrophic lakes in southeastern Wisconsin to very clear oligo- 
trophic lakes in the northern part of the state, is shown on Map A. 
Eight different ERTS images were required to provide coverage of all 37 
lakes. Secchi depth readings were selected as the ground truth measure 
of lake eutrophication to be correlated with lake exposure calculated 
from the ERTS image. Secchi depths and various other water quality para 
meters are sampled quarterly in these lakes as a part of the DNR Lake Water 
Quality Monitoring Program. 
The sampling date of each of these lakes was within 25 days of the 
ERTS overflight date.* In addition, each lake 1) had no tannin coloring, 
2) was at least 20 feet deep to minimize bottom interference, 3) was large 
enough to insure that the measurement spot of the microdensitometer was 
wholly within the lake, and 4) was not obscured at all by clouds or 
atmospheric haze. 
A Gamma Scientific spot microdensitometer equipped with a digital 
readout photomultiplier-picoammeter combination was used for the measure 
ment of the transmitted intensity of light through the film. A measurement 
spot size diameter of 50 microns, which corresponds to 550 feet on the 
ground, was selected for the analysis of lake imagery. This is large 
enough to average across several of the pixels or resolution cells of 
the ERTS multispectral scanner which are about 200 feet across on the 
ground. A one millimeter measurement spot size was used for densitizing 
the film wedges on each of the ERTS images. 
The raw current readings output from the densitometer were used to 
calculate the transmittance of light through the transparency for the lake 
of interest. On any one frame, lake image transmittance might be expected 
to correlate with secchi depth. However, the transmittances of lakes on 
different frames are not comparable because of photographic processing 
differences. These processing differences can be normalized by using the 
film wedges provided on each frame to calculate the relative exposure of 
each lake. Relative exposure is proportional to the light energy hitting 
the ERTS multispectral scanners and is comparable from frame to frame. 
Each step on the film wedge was exposed during processing by an amount of 
For 30 of the lakes the sampling date was within 10 days of the ERTS 
overflight. The remaining lakes (sampled within 11-25 days) were 
included because they were all known to be oligotrophic and were not 
expected to show much variability in algal turbidity over time.