| Spatial variation characteristics of soil particle composition (SPC), bulk density (BD), soil nutrients and CEC in the Huang-Huai-Hai Plain (HHHP) were studied by coupling Geographic Information System (GIS) with geostatistics in this paper. A conjunctive simulation model incoporating soil water, heat, nitrogen behaviors and crop growth was established and validated using field experiment data, and then coupled with GIS to estimate soil water and nitrogen dynamics at regional scale. The regional distribution of water and nitrogen use efficiency (NUE), nitrogen loss under summer maize-winter wheat rotation system were studied under the meteorological and rural socio-economical condition of HHHP, and the optimal water and nitrogen treatments were determined based on the simulation results. The main results were as follows:SPC, BD of each soil layers and soil nutrients, CEC of 0-20 cm soil layer were varied spatially with regional features. Their trend effects and anisotropy were obvious. The results indicated that the analysis of trend and anisotropy was necessary and suitable to be involved in the study on semivariogram of soil properties at regional scale.Stochastic variability for soil properties is beyond 0.5. Spatial distribution of soil properties distributed as stripe or lump patterns.Pedo-Transfer Functions (PTFs) models using Back-Propagation (BP) neural network were establishedto predict soil hydraulic characteristic parameters. The predicted values agreed with the observed ones fairly well, and the BP neural network model had rather high prediction precision. The T-test result showed that with input of soil texture, it can satisfy the simulation requirements of soil hydraulic characteristic parameters at regional scale..A combined soil water-heat-nitrogen and crop growth simulation model was established and validated based on the field data. It's found that the simulated values on soil water and nitrogen content fit well with the observed ones. Coupling with GIS by using VBA environment embedded within ArcGIS8.1 and developing the relevant sharing database, the combined model could be used for water-heat-nitrogen simulation at regional level.The GIS-based model was used for regional simulation on summer maize-winter wheat rotation system in 1999-2000. The multivariate stepwise regression results show that in summer maize-growing season, nitrogen leaching was significantly positive correlation with nitrogen fertilizer amount (NFA), saturated hydraulic conductivity (Ks), Active accumulated temperature (ACT ), rainfall and irrigation amount (RIA), water use efficiency (WUE) was significantly positive correlated with NFA and significantly negative correlated with ACT, nitrogen use efficiency (NUE) was significantly positivecorrelated with RIA and significantly negative relative with NFA and soil organic matter (SOM). In winter wheat-growing season, Nitrogen leaching was significantly positive correlated with NFA, Ks and RIA, WUE was significantly positive correlated with NFA and sunshine hours (SH) and significantly negative correlated with RIA and Ks. NUE was significantly positive correlated with RIA and SH and significantly negative correlated with NFA and Ks. WUE, NUE and Nitrogen leaching.Due to the variations in biophysical conditions and farming practices, the inter-physiognomy variations on WUE, NUE as well as soil nitrogen leaching has exhibited significant differences.A multiple goal hybrid programming model was established based on simulation of regional soil water and nitrogen behaviors. Its programming variables were irrigation and nitrogen treatment. The programming goals were annual grain yield, WUE, NUE and Nitrogen leaching. The result shows that under meteorological condition in 1999-2000, in rainfall-scarce regions in Beijing, Tianjin, Hebei, central and northwestern Shandong, the ideal suitable water and nitrogen treatment were I2F2 (320 mm, 280 kgN.hm-2) and I1F2 (220 mm, 280 kgN.hm-2). In other regions where rainfall was sufficient relatively, the optimal water and nitrogen treatment we...
|
PDF Full Text Request