Seasonal frozen ground freeze-thaw cycles in cold regions are an essential indicator of climate change, infrastructure, and ecosystems in the near-surface critical zone (CZ). As a non-invasive geophysical method, the ambient noise seismic method estimates the relative velocity variations (dv/v) based on coda waves or ballistic waves, providing new insights into the seasonal frozen ground changes in the soil properties and hydrology data, such as soil moisture content (SMC), temperature, and groundwater level. Due to the dv/v lack of accurate depth information and average over tens of days at low frequencies, it is challenging to provide the needed temporal-spatial resolution for the micrometer-level frozen ground variation. In this work, we combine the 1D linear three-component seismic array and hydrological sensor to conduct seasonal frozen ground freeze-thaw monitoring experiments. Besides the conventional dv/v information, we calculate surface-wave (SW) dispersion curve variations (dc/c), which are more sensitive to SMC and can characterize the daily air temperature variations. Meanwhile, the horizontal-to-vertical spectral ratio (HVSR) amplitude and seismic attenuation also show highly consistent changes to the freeze-thaw processes. This work demonstrates that the different ambient noise seismic information (dc/c, HVSR, and attenuation) provide robust observations for hydrogeological monitoring, such as air temperature, SMC, and groundwater level changes during seasonal freeze-thaw processes.

Heat flow is a geothermal parameter for indicating the heat sources distribution and evaluating geothermal reservoirs. Only 1230 heat flow points are distributed unevenly in China, mainly concentrated in the high-temperature geothermal areas and the southeast regions. The Songliao Basin is a potential geothermal field in China. Still, only 20 measurement points are known, making it difficult to evaluate the geothermal genetic mechanism. Sparse data interpolation using deep learning methods have high accuracy and are widely used in fields such as image processing. In this work, we propose a deep neural network for predicting heat flow in the Songliao Basin. More than 4,000 global heat flow and 23 geological and geophysical parameters are used as reference constraints for training. The uncertainty error of the prediction is estimated based on the correlation and distance-based generalized sensitivity analysis. The results show that the maximum heat flow is 85 mW/m2, the average is 67.1 mW/m2, and the error with the measured data is 10.64%. The previous geophysical and geological interpretation results indicate that the heat flow is higher in the west and lower in the east, with high anomalies in the central region, which may be related to the uplift of the deep mantle and the depression of the shallow low-velocity sedimentary layer. Some high-temperature melt bodies are in the deep layers, forming the current potential geothermal field. The measured data validates that the DNN is an effective method for predicting regional-scale heat flow, providing reliable heat source information for evaluating geothermal resources.

One of the scientific payloads of Chang’E-5 (CE-5), i.e., the lunar penetrating array radar (LRPR), will carry out the in situ exploration of the regolith structure and guide the drilling sampling process. To evaluate the performance of the LRPR system, we present a multifrequency full-waveform inversion (FWI) with the total variation (TV) regularization constraint using simulated CE-5 LRPR data for imaging the regolith structure and estimating the physical parameters (permittivity and conductivity). The multifrequency FWI strategy is used to improve the inversion resolution, which updates the low-frequency gradient for the deep region and then increases the frequency range to update the shallow region. The TV regular-ization constraint not only reduces the gradient noise but also improves the inversion accuracy of local structures. To evaluate the actual LRPR measurement scenario, we use the actual source wavelet obtained from the LRPR instrument prototype in a ground lab to replace the theoretical Ricker wavelet. The actual source includes the effect of antenna radiation patterns and clutter noise from the metallic lander. Two typical heterogeneous lunar regolith model tests demonstrate that the proposed FWI scheme effectively reduces the lander metal impact, provides a reliable way to estimate the lunar regolith physical parameters and image the subsurface structures.

Revealing the dynamics of groundwater movement in the vadose zone is crucial to groundwater management and artificial recharge. The traditional hydrographic surveys are usually indirect, costly, and infrequent. In this study, the groundwater flow characterization of a pumping experiment in Jilin University campus was monitored through combining time-lapse electrical resistivity tomography (ERT) and self-potential (SP) data tomography. The ERT datum depicts the spatial distribution of resistivity, which is related to the dynamics of soil moisture content during the pumping process. We are able to correlate hydraulic heads and SP signal during a decline and recovery groundwater level period leading to interesting perspectives in understanding the dynamics of complex groundwater flow. The SP method provides a direct way to estimate the potential field distribution, which can be further used to invert soil permeability. A total of 24 hours of time-lapse geophysical surveys revealed a significant increase in resistivity but a decrease in permeability during water pumping and groundwater recharge, representing the process of groundwater decline and recovery. Results derived from time-lapse geophysical surveys matched well with in situ monitoring of the groundwater level. The study demonstrates that the combined ERT and SP data can provide a direct and reliable way to monitor groundwater flow or other time-lapse hydrogeological surveys.

Revealing the dynamics of groundwater movement in the vadose zone is crucial to groundwater management and artificial recharge. In this study, the groundwater flow characterization of the pumping process is monitored by the joint time-lapse electrical resistivity tomography (ERT) and self-potential (SP) data tomography. The ERT data invert the resistivity distribution, which relates to the variation of soil moisture content during the pumping process. Base on the groundwater motion feature, the SP data provide a direct way to invert the current density and estimate the permeability. A total of 24 hours of time-lapse surveys show a significant increase and decrease in resistivity and permeability during water pumping and groundwater recharge, which suggests groundwater decline and recovery process. These results have an excellent agreement with the groundwater level monitoring result. Combining ERT and SP data can provide a reliable way in groundwater or other hydrogeological surveys.