Potential impact of controlled drainage in Indiana watersheds

Yinghui Sui, Purdue University

Abstract

Due to naturally high groundwater tables, many soils in the Midwest need artificial drainage for economical crop production. However, nitrate carried in drainage water can lead to water quality problems. Controlled drainage is a strategy to reduce the nitrate loads and maintain adequate drainage from crop production by regulating drainage depth with a water-level control device at the drainage outlet. This dissertation addresses nitrate loads from subsurface drainage and the potential benefit of controlled drainage at the watershed scale. Two methods were applied to estimate the contribution of tile drains to watershed nitrate load: a statistical analysis of many watersheds and a simulation of water flow and chemical processes for a particular watershed. Watershed boundaries were delineated for 29 gaging stations in Indiana, where nitrate and flow have been monitored, and the percentage of each watershed that is drained by subsurface tile drains was estimated from soil, slope, and land use characteristics. Statistical models were developed to relate the percentage of the watershed with subsurface drains and the non-point source nitrate loads, with and without precipitation as an additional explanatory variable. Significant relationships were found for annual models both with and without precipitation, from February to June for monthly models without precipitation and from January to July for monthly models with precipitation. Expected annual and monthly non-point source nitrate loads were calculated for various percentages of drained area in Indiana watersheds based on the models with precipitation. Predicted annual non-point source nitrate load ranged from 383 kg/yr/km2 for a watershed with 0% drained area to 2750 kg/yr/km2 for a watershed with 100% drained area, which suggested that about 86% of non-point source nitrate coming from a 100% drained watershed would flow through tile drains. SWAT2005, with modified tile drain components, was used to simulate nitrate loss through tile drains in a heavily tile drained watershed in Indiana, the Sugar Creek watershed. Nitrate loads from surface flow, lateral flow, percolation and tile drains were estimated. The median percentage of monthly nitrate input to Sugar Creek that came from tile drains ranged from 0% to 37% over the 12 months, and was more than 30% from April to June. These estimates of nitrate from tile drains were used to predict the potential effects of controlled drainage at the watershed scale. Previous field and plot scale research in the Midwest suggested that the likely percentage reduction was 25% to 47% in Indiana. For 29 watersheds in Indiana, areas with high and medium potential for controlled drainage were estimated based on appropriate crop, soil types and topography. If controlled drainage was applied in all fields with high potential, the watershed-scale percentage of reduction could reach 8.7-16.4%. If areas with medium potential also had controlled drainage installed, the reduction could reach to 14.3-26.9%.

Degree

Ph.D.

Advisors

West, Purdue University.

Subject Area

Environmental engineering

COinS