APPLICATION OF REMOTE SENSING FOR ADJUSTMENT AND CALIBRATION OF WATER QUALITY MODEL - Integrating GIS, Remote Sensing, and Mathematical Modelling for Surface Water Quality Management - page 178

Figure (8-23) illustrates the relationship between the three data sets; a positive trend is observed. It is
concluded that the model is underestimating the measured values and the MODIS extracted values.
Calculating the RMSE between the measured and MODIS TSM values is 2 mg/l, and between the
measured and modelled TSM values is 5.12 mg/l. The developed TSM relation could be used to
develop a TSM time series by applying it to the set of images covering the modelling period. It is
noted that the model is under predicting the TSM concentrations compared to the measured and
extracted values, but it is still considered within an acceptable range of RMSE. Therefore the
developed time series could be used to calibrate the model through intensifying the model forcing
data. The selected locations for developing the TSM time series are the exits of Edko and Barseek
drains into the lake. To develop the time series from the images, first the algorithm is applied to all the
images, and then the layer stacking tool under ENVI is used to form a stack of layered images
representing the temporal sequence of the images. A vertical spectral profile is drawn intersecting
these grouped images at the selected pixel locations to represent the time series. Figures (8-24) s hows
the developed time series at both locations.

Time Series -Edko Drain Exit

Time Series -Barseek Drain Exit

60.00

45.00

40.00

50.00

35.00

30.00 40.00

25.00

30.00

20.00

15.00 20.00

10.00

10.00

5.00

0.00 0.00

TSM (mg/l) TSM (mg/l)

Figure (8-24): TSM concentrations time series from MODIS TSM maps at Edko and Barseek drains exits into
the lake

8.5.5. Calibration of Model Using Developed MODIS TSM Concentration Maps

The mathematical model was calibrated for TSM using the developed time series as a new input
dataset at the exit points of both drains entering the lake as shown in Figure (8-25) . The new predicted
values are changed slightly are due to the effect of the accuracy of the developed algorithm and the
site characteristics which need detailed information on inherent optical properties (IOP's) of the water
body under investigation. Also some spectral measurements taken during water quality sampling
reflect the mixed properties of the water body since the shallow lake is dominated by the presence of
the submerged vegetation Potamogeton pectinatus, and this can affect the measured concentrations of
TSM.

Figure (8-25): Model calibrated values, image extracted values and the corresponding field measured TSM
concentrations on 28 th June