Analyzing representativeness of monitoring units for pollutant removal in agricultural drainage ditch-pond systems

Abstract: In order to decide the pollutant removal ability of ditches and ponds in agricultural regions, it is often required to monitor the flow and pollutant dynamics. Differing from lab and pilot study settings in which the inflow and outflow can be managed and measured, the field ditches and ponds may not have single inlet and outlet clearly defined. Since it is difficult to monitor all ditch-pond segments due to technical and labor costs, a monitoring unit (MU) covering a fraction of the system may be used to deduce the system performance. To obtain a reasonable system estimation, a pre-requisite of MU is that it has to be representative, otherwise the overall performance of the system deduced from a partial MU will be erroneous. The mixed distribution of drainage ditches and ponds with crop fields causes the variation in pollutant removal among different segments, and this can affect the representativeness of MUs consisting of one or several segments. It is affected by the distribution, area and hydraulic connections of ditches and ponds, as well as pollutant degradation properties. This paper is based on a case study in Zhaoguan Irrigation District along the Grand Canal in the Jiangdu District of Yangzhou, China (119°25'E, 32°22'N). In the study area, the drainage ditch-pond system covers an area of 0.80 hm2, including 10 smaller field ditches covering an area of 0.09 hm2, 6 bigger branch ditches and 2 ponds. The total crop field is 5.61 hm2; and the areal ratio of ditch-pond to crop field was 14.3%. Previous investigation has shown that the pollutant removal in such field conditions was affected by ditch-pond distribution, area and hydraulic relationship. This paper investigated the representativeness of selected MUs and the factors for 2 most commonly used system evaluation indices: pollutant removal rate and intensity. Methods to evaluate and improve the representativeness of different MU compositions were proposed for pollutants with different degradation properties. The results indicated that the majority of segments could not represent the system adequately; while a particular monitoring unit could have excellent representativeness of one index, it could have rather poor representativeness of the other index. The overall segment removal intensity ratio was larger than 1, while the removal rate ratio was smaller than 1. For an intermediate pollutant degradation (degradation coefficient was 0.05 m/d), the removal intensity ratio was 1.54, and the removal rate ratio was 0.59, suggesting that using a single segment as MU will result in more than 50% overestimation of pollutant removal ability, while more than 40% underestimation of pollutant removal rate. The representativeness was affected by pollutant degradation properties and hydraulic retention time. The ratio of removal intensity had positive correlation with the hydraulic retention time while the ratio of removal rate had negative correlation with the hydraulic retention time. When degradation coefficient was between 0.01 and 0.10 m/d, the removal intensity ratio and the removal rate ratio for Drainage path 1 were 1.28-1.14 and 1.21-1.16, respectively. As for Drainage path 3, the 2 values were 0.84-0.94 and 1.27-1.33, respectively. Considering the amount of monitoring work, Drainage path 3 (Branch ditch 5) was better selected as the monitoring unit. The selection of MUs will be more forgiving for lower degradable pollutants, and bigger segments are generally better selection than smaller ones. For pollutant removal intensity, the selection of MUs can be more forgiving for low degradable pollutants; and for higher degradable pollutants, smaller segments should be avoided. For pollutant removal rates, smaller segments should be avoided, and it was recommended that the segments located at the middle stream should be selected. For the conditions when drainage paths, fields and pond/ditch segments were distributed unevenly, drainage paths with large volume should be considered as the monitoring unit and small drainage paths should be avoided. For the areas with multiple drainage paths, the segments with larger flow should be selected as the MUs, and the segments with disproportional ratios of ditch-pond to crop field should be avoided.