融入宽范围物性约束策略的大地电磁和重力的Gramian约束联合反演

Magnetotelluric and gravity joint inversion using Gramian constraints integrated with strategy of wide-range petrophysical constraints

宽范围物性约束方式可以促进物性参数在一定范围内实现耦合,具有一定的容错性,实现简单.但目前该方式只在全局优化算法联合反演中得以实现,如何将其推广至梯度优化算法联合反演至今没有明确的策略.此外,Gramian约束虽然对先验信息依赖低,但是对明确的物性关联信息利用率低.基于此,本文提出了适用于梯度优化算法的宽范围物性约束的策略,即"岩石物性关联+范围约束+耦合项",并实现了基于宽范围物性约束策略的大地电磁测深(MT)和重力共轭梯度Gramian约束联合反演新算法.模型试验表明,本文提出的算法可以提高先验物性关联信息的利用率,且有效限定物性耦合的范围.我们将本文方法应用于安徽省茶亭铜金矿床实际资料处理之中,联合反演结果清晰地揭示了铜金矿床的空间分布.

Integration of wide-range petrophysical constraints can facilitate the coupling of petrophysical parameters within a certain range,has certain fault tolerance,and is easy to implement.Nevertheless,this approach is currently only implemented in global optimization algorithms.The strategy for extending the wide-range petrophysical constraints to gradient optimization algorithms remains unresolved.In addition,although Gramian constraints depend on prior information to a small extent,the utilization of explicit related information of petrophysical parameters is low.Thus,we propose the strategy of wide-range petrophysical constraints suitable for gradient optimization algorithms,namely,"petrophysical property correlation plus range constraints and coupling terms".We also realize the joint inversion of magnetotelluric(MT)sounding and gravity based on the new strategy using conjugate gradient algorithm combined with Gramian constraints.Model tests show that the strategy can improve the utilization of prior information,and effectively limit the range of the coupling of petrophysical parameters.This method is also applied to the Chating copper-gold deposit in Anhui province and the joint inversion results clearly reveal the spatial distribution of copper-gold deposit.