COS 14-2 - Linking land use metrics to water quality of drinking water reservoirs at multiple scales

Tuesday, August 13, 2019: 8:20 AM
M101/102, Kentucky International Convention Center
Lilian Ding, Xin Chen, Jianjun Tang and Liangliang Hu, College of Life Sciences, Zhejiang University, Hangzhou, China
Background/Question/Methods

Drinking water reservoirs play an important role in providing water quantity and quality for human societies. Understanding the associations between land use metrics and water quality of reservoirs would help water quality security by a better land use planning. In this study, we examined whether and how land use affects water quality of reservoirs by analyzing 61 reservoirs covering most types of the reservoirs in southern China over the period of 2015-2016. We first analyzed the spatial and temporal distributions of water quality parameters including pH, total nitrogen (TN), total phosphorus (TP), chemical oxygen demand (CODMn), and ammonia nitrogen (NH3-N). Land use metrics including components and configuration were also analyzed. We then calculated the spearman correlation coefficient between water quality parameters and land use components. We also compared the effect of multiple land use configuration indicating by landscape pattern indexes on water quality using redundancy analysis (RDA). R software was used to generate heat map, and cluster map of the correlation coefficients between water quality parameters and land use components.

Results/Conclusions

Water quality parameters of the reservoirs varied significantly at different altitudes in our study region. Land use components and land use intensity (LUI) were also varied among the reservoirs. The percentage of construction land had significantly positive correlation with CODMn and NH3-N, and had significantly negative correlation with DO. The percentage of forest and grassland had significantly negative correlation with CODMn, TN and NH3-N. For the percentage of cropland, however, no significant correlations were found between land use components and the water quality parameters except for the CODMn. The LUI positively correlated with CODMn, NH3-N, TP and TN. In the cluster, the grassland and forest were clearly distinguished from the cropland, construction land and LUI. RDA analysis also showed that the percent of landscape of construction land was positively correlated with DO, CODMn, NH3-N, TP and TN. RDA further showed that when distance of the buffer from water of the reservoirs was 1500m that high percentage of construction land was found, the effects of land use configuration on water parameters were strongest in contrast to the scales of 100m, 200m, and 500m of surrounding buffer, and the catchment scale. These results suggested that an effective land use planning at the scale of 1500m could help to improve water quality. This study highlights the importance of land use metrics for the water quality of drinking water reservoirs.