76 
dissolved oxygen. When interpolating water quality mapping data collected 
throughout the day, this variability presents a potential problem that is best illus¬ 
trated by a map. Figure VII-8 shows that data collected early in the morning on one 
side of the Severn River in Maryland is substantially lower than data collected later 
in the day on the other side. If these measures were interpolated, it would appear that 
one side of the river is faring more poorly than the other when, in fact, the dichotomy 
merely represents a temporal artifact. 
To produce a more representative spatial interpolation of surface dissolved oxygen 
data, estimating the diel dissolved oxygen trend from continuous monitoring instru¬ 
ments and using that trend estimate to adjust the Dataflow dissolved oxygen may 
prove more feasible. The University of Maryland investigated this procedure by 
comparing data from a nearshore continuous meter with those from a mid-channel 
continuous buoy. They found that the dissolved oxygen in the two locations 
responded differently to the local habitats and that nearshore dissolved oxygen 
dropped at night and the mid-channel dissolved oxygen was highly variable, often 
exceeding dissolved oxygen saturation during the day. Although the adjustment 
procedure improved the data set, the prediction error was high. Further research is 
needed to integrate the spatial and temporal monitoring data. 
Figure VII-8. Illustration of rising dissolved oxygen concentrations during the day 
(June 28, 2001) in the Severn River, Maryland. . 
chapter vii 
Shallow water Monitoring and Application for Criteria Assessment 
