Keywords
monitoring system, water quality indicator, chemical load, monitoring design, uncertainty
Start Date
1-7-2008 12:00 AM
Abstract
Efficient assessment of water quality requires an integrated approach incorporating such tools as sampling programs supplying data for statistically valid assessment along with watershed and water quality models which are used to transform the collected data into information. It is important to quantify and possibly to reduce the uncertainty of the information supplied by a monitoring system. The paper describes the algorithm for an efficient temporal monitoring design for data collection at a given site of a fixed station tiered monitoring system. The algorithm takes into account features of models used to obtain estimates of investigated water quality indicators. Efficiency of a monitoring design is considered under the assumption that the cost of a sampling program is a monotonously increasing function of a number of observations. Application of the proposed algorithm to various water quality ingredients reveals a wide range of sampling frequencies needed for estimation of water quality ingredient load with a desired level of uncertainty. Monitoring design which is common for all observed water ingredients at a given site and supporting evaluation of their annual loads with the same level of uncertainty can hardly be attained due to financial and technical constraints. For tiered monitoring systems, different allowable levels of uncertainty can be recommended for water ingredients according to their importance for a particular site.
Role of Models for Building an Efficient Monitoring Design
Efficient assessment of water quality requires an integrated approach incorporating such tools as sampling programs supplying data for statistically valid assessment along with watershed and water quality models which are used to transform the collected data into information. It is important to quantify and possibly to reduce the uncertainty of the information supplied by a monitoring system. The paper describes the algorithm for an efficient temporal monitoring design for data collection at a given site of a fixed station tiered monitoring system. The algorithm takes into account features of models used to obtain estimates of investigated water quality indicators. Efficiency of a monitoring design is considered under the assumption that the cost of a sampling program is a monotonously increasing function of a number of observations. Application of the proposed algorithm to various water quality ingredients reveals a wide range of sampling frequencies needed for estimation of water quality ingredient load with a desired level of uncertainty. Monitoring design which is common for all observed water ingredients at a given site and supporting evaluation of their annual loads with the same level of uncertainty can hardly be attained due to financial and technical constraints. For tiered monitoring systems, different allowable levels of uncertainty can be recommended for water ingredients according to their importance for a particular site.
