Varied groundwater pumping-induced subsidence in stratified soil condition using Influence Function and GIS-derived data

The excessive groundwater pumping and different hydraulic and geotechnical soil properties due to the stratified soil around pumping wells are of the most important factors on the land subsidence. Some consequences of land subsidence include the formation of cracks on the ground surface, destruction of buildings, damage to water conveyance pipelines embeded in the ground, the change in the longitudinal river steepness and vulnerability of residential areas to flood events due to lower soil permeability caused by the land subsidene. Because of the widely use of GIS to the evaluation of earth properties, many researches have been conducted so far on the application of GIS to land subsidence analysis. In this regard, a number of GIS-based researches focus on the application of data derived from In-sar radar to monitor the land subsidence without any geotechnical analysis. Due to the necessity of geotechnical analyses to more accurately approximate the land subsidence, the researchers proposed a method based on the influence function in order to relate the variation of phreatic groundwater surface around the pumping wells to the deformation of ground surface causing the land subsidence. Despite the incorporation of soil stratification effect, the inattention to variations of phreatic surface in variable pumping conditions from multiple wells and the mutual effects of wells on each other is accounted as a weakness of the proposed method which is addressed in the present study.

In the present research, a practical computer program (so-called CALS-SVD) was developed which initially imports the physical and hydraulic properties of an aquifer from GIS including the longitude and lattitude of wells, the specific storage coefficient of the soil, hydraulic conductivity, initial void ratio of the soil, soil compression index, the wet and saturated specific weights of the soil. Then it applies a new method integrating Influence Function with equations of varied pumping-induced influence radius of groundwater surface to estimate the subsidence (from multiple wells in a given time). Finally, the results of computations are added as new information layers to the imported GIS file. The developed computer model was applied to a case study in west of Guilan province located in the road connecting Saravan to Fouman. The applied approach in the studied area was to divide it into two regions. Then, 6 scenarios were considered in terms of reported pumping rates (ranging 30 to 120 m3/d). The maximum reported pumping rate in the region (120 m3/d) and the half and quarter of the maximum pumping discharge were applied to the wells located in two mentioned regions in order to model the land subsidence in the entire studied area. Taking into account the overlap of pumping wells, one of the wells located in the middle of others was selected as the origin and the effect of all the wells on each other were determined based on the relative distance derived from their longitudes and lattitudes.

The scenario A120B120 incorporating the maximum pumping discharge at the both divisions of the studied area used to validate the model. Validation of the obtained results with the field surveys shows an admissible performance of the presented software. Additionally, the results showed that the pumping-induced subsidence up to 30 cm around the wells leads to the expansion of subsidence area in the region. But a greater subsidence depth has negligible influence on the subsidence area, and it will be limited to 70 percent of the total surface area of the region. Meanwhile, the higher the pumping rate, the greater will be the influence radius. But the rate of increase in influence radius decreases for higher pumping values. Thus the average influence function will be limited to 105 meter at the studied area. According to the results, there is a linear relationship with an admissible correlation coefficient between the average groundwater pumping rate (in a 10 year period) and the maximum subsidence depth. Moreover, the maximum subsidence depth is highly affected by the stratification of the soil. Whereas the subsidence increase rate with pumping is mostly affected by the percentage of increase in pumping. Finally, the application of presented subsidence model in the studied regions indicates lower magnitude and rate of subsidence in a 10 year period in comparison with the results obtained from Damghan in a 5 year period.