Ground truth data for chlorophyll a in vivo concentrations, salinity, and 

 algal species identification were supplied by the Virginia Institute of Marine 

 Sciences (VIMS), and water temperature and attenuation coefficients were mea- 

 sured for each test. The results from that test are shown in figure 10 

 (ref. 8). Because of ambiguous data obtained at the laser wavelengths of 

 598 nm and 61 8 nm, only results obtained at the laser wavelengths of 454 nm 

 and 539 nm, which corresponded to optimum excitation of green and golden-brown 

 algae, were used to determine the total chlorophyll a concentration given in 

 the figure. The calculation was made according to a matrix equation (ref. 13) 

 which was similar to equation (11). However, the constant in the power- 

 received equation in reference 13 is a factor of 2 larger than equation (8), 

 and reference 13 defines Yf and Yi as the single-scattering attenuation 

 coefficients oif and a^. Since Yf and Yi> as defined previously, are 

 always less than a.f and a^, the value of P r (X^) would be smaller because 

 of the single-scattering attenuation coefficient assumption. Thus the errors 

 in the analysis of ALOPE data were somewhat offsetting, and the results which 

 were obtained were encouraging enough so that the instrument was subsequently 

 flown over the lower end of the James River in Virginia. Figure 11 shows the 

 flight path of the helicopter on the 138-km flight. The distance from the 

 laser system to the water was approximately 100 m, and, as a result of a 

 2-second interval between each of the four laser firings, one complete data 

 set was obtained over each 270 meters. The averages of the remotely sensed 

 chlorophyll a concentrations, calculated from the power-received equation 

 in references 8 and 14 were obtained for each leg of the flight and are pre- 

 sented only for information in figures 11 and 12. (The error bars represent 

 differences between outgoing and incoming legs of the flight.) Note the 

 2TTm^ difference between equation (8) and the power-received equation in 

 references 8 and 14 and also note attenuation coefficient definition differ- 

 ences. In addition, it should be mentioned that the data obtained with the 

 rhodamine 6G laser, which operates at 598 nm, were not used because of filter 

 blocking problems and because ground truth measurements indicated that there 

 were no red algae present in the water. The fluorescence cross sections given 

 in reference 14 were used in the data reduction, and attenuation coefficients 

 were obtained from in situ measurements over the flight path of the helicopter. 

 No comparison of these results with in situ measurements of chlorophyll a con- 

 centrations was reported. The calculated concentrations of chlorophyll a in 

 each algal color group can only be expected to predict the actual concentra- 

 tions when the proper relationship, as given in equation (8), is used to relate 

 the fluorescence power received by the laser fluorosensor system to the chloro- 

 phyll a concentration in the water. 



ERROR ANALYSIS OF LASER FLUOROSENSOR SYSTEMS 



Single-Wavelength Systems 



The chlorophyll a in vivo concentration is determined for a single- 

 wavelength laser fluorosensor system by equation (7). To find the statistical 

 variance of the chlorophyll a in vivo concentration, it is first assumed that 

 P , P r , (yj, + Yf)> and a are independent random variables and the other 

 parameters in the equation are constants. The general variance equation, which 

 is based upon propagation of errors by least squares (ref. 24), is given by 



14 



