勘探深度较大的瞬变电磁法（transient electromagnetic method, TEM）的有效数据处理技术，是近年来大深度、高分辨、高效率地下构造和资源探查应用的一个重要方向。提出了电磁数据横向约束反演（lateral constraint inversion, LCI）的有限内存拟牛顿算法，并在鄂尔多斯盆地新街一井矿区深埋藏煤层的勘探数据处理中应用，取得良好效果。与常用的一维反演方法相比，本方法具有显著的高分辨、高可靠优势。也为面积性电磁勘探数据处理提供了有效的理论和实际方法技术。结合地质和钻探资料进行的TEM数据横向约束反演处理解释，对厚度几十米、深度上千米的沉积地层和煤层进行了可靠的识别和划分，大致圈定了隐伏可采煤层赋存范围、底板埋深，以及白垩系志丹群、侏罗系直罗组岩层和可采煤层及其含水区域的分布范围，为探采安全生产提供了重要依据。

Development of effective data processing techniques for transient electromagnetic method (TEM) has become an important area in recent years for large-depth, high-resolution, and high-efficiency exploration of underground structures and resources. The finite memory quasi-Newton method (L-BFGS) for three-dimensional lateral constraint inversion(LCI) in TEM data processing was proposed, and satisfied results in the exploration for deeply buried coal seams in the Xinjie Yijing area of Ordos Basin, China has been obtained. Significant advantages of high resolution and reliability of this method was demonstrated by comparison with current one-dimensional methods. Thus, an effective theoretical and practical methodological techniques for the processing of areal electromagnetic exploration data was also provided. With integration of geological and drilling data, the sedimentary strata and coal seams with thicknesses of several tens of meters and depths of over a thousand meters were reliably identified and delineated by the proposed TEM-LCI method. The distribution range, bottom depth of concealed recoverable coal seams, as well as the distribution range of the Cretaceous Zhidan Group, Jurassic Zhiluo Group strata, coal seams, and water-bearing areas of resistivity anomalies within these strata and coal seams have been roughly outlined. Strong support for coal mining and safe production is then provided with the results of this study.