Background
The Minqin Oasis, located in northwestern China, was once a lush and fertile agricultural region that thrived in the midst of the Badanjilin and Tengeli deserts (Figure 1). Fed by the lower reach of the great Shiyang River (Photos 1 and 2), and sustained by a shallow regional water table, the oasis is the ancestral homeland for thousands of farmers, supporting a variety of crops, from cotton to sunflower (Photos 3 and 4). However, because of increasingly higher water extractions along the Shiyang River, and excessive groundwater pumping within the region (Photo 5), the water table has declined significantly over the last several decades, violating the once dynamic equilibrium between desert and oasis. As a recent June, 2006 New York Times front page article reported: “An ever-rising tide of sand has claimed grasslands, ponds, lakes and forests, swallowed whole villages and forced tens of thousands of people to flee as it surges south and threatens to render this ancient Silk Road greenbelt uninhabitable” (Photos 6, 7, and 8). The severe consequences of this regional desertification extends far beyond the greenbelt, as the Minqin Oasis has become the single largest source of dust storms in Asia, even impacting the City of Beijing, located far away on the eastern coast of China.
In addition to drawing great international attention, this environmental disaster has galvanized the government of China, which is committed to halting the desert’s migration and destruction of the oasis through establishment of a national “937 Project.” The Prime Minister Wen Jiabo has traveled to the region to pledge his support, and researchers from around the country are studying different components of the problem in great detail. The ultimate goal is to find an optimal balance between agricultural/ economic viability for the region and environmental protection and sustainability of the oasis.
Figure 1. Map depicting the Minqin oasis, the lower reach of Shiyang River, and the Hongyashan reservoir (Courtesy of: Huo Zailin, Ph.D. Candidate and Professor Shaoyuan Feng, College of Water Conservancy and Civil Engineering, China Agricultural University).
 |
 |
 |
|||
|
|
|
|||
 |
 |
 |
|||
|
|
|
Research Objectives
Mr. Huo Zailin, a Ph.D. candidate at the China Agricultural University, College of Water Conservancy and Civil Engineering, under the direction of his graduate advisor, Professor Shaoyuan Feng, has developed highly accurate ANN groundwater simulation models for the oasis. Trained and validated with historical data spanning 1980 through 1997, the ANN models accurately simulate regional mean monthly transient groundwater levels in response to time varying climate, hydrologic, and agricultural conditions. Figure 2 depicts ANN predicted versus observed groundwater levels for validation data corresponding to the Xinhe region. In addition to serving as a highly accurate simulation and forecasting tool, the ANN models also revealed the dynamic inter-relationships between groundwater levels in the oasis and numerous factors, including surface water levels in the Hongyashan reservoir, agricultural extractions, and climate conditions. The ANN models have been used to project future groundwater levels under a variety of possible climate and agricultural conditions, as shown by Figure 3 for a representative case. A scientific paper overviewing this research has been submitted by the University researchers to the Journal of Hydrology.
Mr. Zailin and Professor Feng are now developing an ANN-based surface water model for the Shiyang River, which will be coupled with the ANN-based groundwater models for the Minqin Oasis, constituting a fully integrated groundwater-surface water model for the region. In conjunction with NOAH, multiobjective optimization will be performed with the models, providing the government with its optimal set of alternatives for balancing agricultural production with protection of the oasis through improved water resources management. That is, the optimal allocation of scarce water resources, among a disparity of regional users, including agricultural, industrial, and urban, will be identified in accordance with the appropriate trade-off between economic and environmental objectives.
Ultimately, the integrated ANN-based groundwater-surface water simulation and management model can be used as a real-time prediction and management tool, optimizing water allocations within the region based upon existing and projected (e.g. weather) conditions.
 |
 |
||
|
|
||
 |
 |
||
|
|