Modeling nitrogen loading in a small watershed in southwest China using a DNDC model with hydrological enhancements

Abstract

Abstract. The degradation of water quality has been observed worldwide, and inputs of nitrogen (N), along with other nutrients, play a key role in the process of contamination. The quantification of N loading from non-point sources at a watershed scale has long been a challenge. Processbased models have been developed to address this problem. Because N loading from non-point sources result from interactions between biogeochemical and hydrological processes, a model framework must include both types of processes if it is to be useful. This paper reports the results of a study in which we integrated two fundamental hydrologic features, the SCS (Soil Conservation Service) curve function and the MUSLE (Modified Universal Soil Loss), into a biogeochemical model, the DNDC. The SCS curve equation and the MUSLE are widely used in hydrological models for calculating surface runoff and soil erosion. Equipped with the new added hydrologic features, DNDC was substantially enhanced with the new capacity of simulating both vertical and horizontal movements of water and N at a watershed scale. A long-term experimental watershed in Southwest China was selected to test the new version of the DNDC. The target watershed's 35.1 ha of territory encompass 19.3 ha of croplands, 11.0 ha of forest lands, 1.1 ha of grassplots, and 3.7 ha of residential areas. An input database containing topographic data, meteorological conditions, soil properties, vegetation information, and management applications was established and linked to the enhanced DNDC. Driven by the input database, the DNDC simulated the surface runoff flow, the subsurface leaching flow, the soil erosion, and the N loadings from the target watershed. The modeled waterCorrespondence to: B. Zhu (bzhu@imde.ac.cn) flow, sediment yield, and N loading from the entire watershed were compared with observations from the watershed and yielded encouraging results. The sources of N loading were identified by using the results of the model. In 2008, the modeled runoff-induced loss of total N from the watershed was 904 kg N yr −1 , of which approximately 67 % came from the croplands. The enhanced DNDC model also estimated the watershed-scale N losses (1391 kg N yr −1 ) from the emissions of the N-containing gases (ammonia, nitrous oxide, nitric oxide, and dinitrogen). Ammonia volatilization (1299 kg N yr −1 ) dominated the gaseous N losses. The study indicated that process-based biogeochemical models such as the DNDC could contribute more effectively to watershed N loading studies if the hydrological components of the models were appropriately enhanced.

Comparison of the DNDC, LandscapeDNDC and IAP-N-GAS models for simulating nitrous oxide and nitric oxide emissions from the winter wheat–summer maize rotation system

“…These results agreed with the range of 7-19% observed at a site in the same region (Cai et al, 2002). At YT, the annual nitrogen losses via NH 3 volatilization given by all of the models accounted for 12-29% of the applied nitrogen, which were in accordance with the value of 24% previously simulated with the modified 92 version of DNDC in the Yanting catchment (Deng et al, 2011a). Compared with the observed NO 3 − leaching from same soil depth, the DNDC95 and LDNDC simulations of the annual loss rates of applied nitrogen in the NO 3 − leaching pathway tended to be lower at YJ (2-4% versus a range of 3-12% in Liu et al, 2012) and higher at HT (4-9% versus 0.3% as reported by Li et al, 2014), whereas the IAP-N-GAS simulations of 15-18% were higher at both sites.…”

Identifying critical nitrogen application rate for maize yield and nitrate leaching in a Haplic Luvisol soil using the DNDC model

“…Some models based on hydrology are commonly used to predict nutrient loadings in watershed scale, such as SWAT (Wang et al, 2014a) and MIKE SHE (Vansteenkiste et al, 2013). There are also several nutrient-loading prediction models, which include components to simulate not only hydrological processes but also nutrient transport and transformation, such as the Denitrification-Decomposition (DNDC) model (Deng et al, 2011) and STICS soil-crop model (Jego et al, 2008).…”

The development of the DNDC plant growth sub-model and the application of DNDC in agriculture: A review

“…For instance, the phenological growth model for crops in DNDC are still in the code but are commented out, and a simpler GDD-based model is employed; the leaf area index becomes an estimated model output and is not involved in crop growth simulation in the model. A total of 63 crop types can now be simulated in cropping systems in the model, including fallow, maize, winter wheat, soybean, legume hay, non-legume hay, spring wheat, sugarcane, barley, oats, alfalfa, sorghum, cotton, rye, vegetables, papaya, potato, beet, paddy rice, and banana Deng et al, 2011a;Li et al, 2012a).…”

“…(6) DNDC also provides tools to quantify the uncertainties for site or regional simulations. With the Monte Carlo approach at the site or grid cell scale, and with the MSF (most sensitive factor) method at regional scale, the uncertainties produced from the model simulations can be quantified (Deng et al, 2011a).…”