This paper investigates the impact of heterogeneity of the transmissivity field on the interpretation of steady-state pumping test data from aquifer systems delimited by constant head boundaries such as aquifers adjacent to lakes or rivers. Spatially variable transmissivity fields are randomly generated and used to simulate the drawdown due to a pumping well located at different distances from a constant head boundary. The steady-state drawdown simulated at different observation wells are then interpreted using the Hantush method (Hantush 1959). The numerical simulations show that, in contrast to the case of infinite aquifer domains, the interpreted transmissivity varies depending on well locations and the separation distance between pumping well and boundary relative to the correlation length. The ensemble-averaged estimated transmissivity varies between the geometric mean and the arithmetic mean, and can even exceed the arithmetic mean in a narrow domain adjacent to the boundary. It approaches the geometric mean of the underlying transmissivity field only if the distance between the pumping well is more than 20 times the characteristic length of the transmissivity field.

中文翻译：

---

恒水头边界非均质含水层抽水试验的解释

本文研究了透射率场的异质性对解释由恒定水头边界（例如与湖泊或河流相邻的含水层）划定的含水层系统的稳态抽水试验数据的影响。空间可变的透射率场是随机生成的，并用于模拟由于泵井位于距恒定水头边界不同距离而导致的水位下降。然后使用 Hantush 方法（Hantush 1959）解释在不同观测井模拟的稳态压降。数值模拟表明，与无限含水层域的情况相比，解释的透射率随井位置和抽水井与边界之间相对于相关长度的分离距离而变化。整体平均估计透射率在几何平均值和算术平均值之间变化，甚至在边界附近的窄域中可能超过算术平均值。只有当抽水井之间的距离大于透射场特征长度的 20 倍时，它才接近下伏透射场的几何平均值。