The dynamics of water flow and sediment transport in river systems play a crucial role in shaping river morphology and the broader watershed management. However, these characteristics are often challenging to measure comprehensively. We studied a low-flow river system (≤0.611 m 3/s) within the boundaries of Yuancun in the Township of Meishui. By synchronously monitoring the microseismic signals generated by the river and the river flow velocity, explored the relationship between these microseismic signals and the river discharge. The signals exhibited a wide frequency range (2–50 Hz). An analysis of the recorded microseismic signals and the flow data revealed an approximate linear relationship between the seismic noise in the 2–10 Hz bandwidth and the river flow. We used a least squares regression model to invert the river flow from the 2–10 Hz microseismic signals and found that the maximum relative error between the inverted flow and the measured values was 10.3%. The results show that even at low flow rates, real-time monitoring of river processes is possible through continuous time-frequency analysis of microseismic signals; this increases the potential for future applications of seismic monitoring in real-time observation of hydrological evolution in river systems.