1. Introduction The thickness of ice layers in reservoirs, rivers, lakes, oceans, and glaciers, as well as the changes in ice formation and melting within the ice layer, have long been research topics in various scientific and engineering fields, including hydrology, geology, meteorology, environmental protection, flood control and disaster reduction, and hydropower engineering, both domestically and internationally. Ice jams in northern China threaten the lives of people along the Yellow River basin and the safety of various hydropower dams. Understanding the patterns of river ice thickness and ice formation and melting is crucial for forecasting and preventing ice jam disasters. In recent years, rising global temperatures and the melting of polar glaciers have become significant factors affecting the human living environment. Changes and impacts on the human living environment caused by global warming have become one of the most pressing issues for humanity in the 21st century. Observing changes in the thickness of glaciers and sea ice in the Arctic and Antarctic can provide the most direct meteorological and hydrological data for analyzing global climate change. Current methods for ice thickness detection include water level measurement, mechanical ice thickness detection, methods utilizing the difference in conductivity between ice and water, electromagnetic induction ice thickness detection, and pulse radar detection. Capacitive ice thickness sensors utilize the difference in dielectric properties between ice and water for ice thickness detection. 2. Capacitance of Air, Ice, and Water as a Perspective of Temperature Change From physics, we know that c = εrs/d, and the capacitance depends on the dielectric constant εr, the plate area s, and the distance d between the plates. For a capacitor with a fixed distance d and plate area s, the capacitance C varies with the dielectric constant εr of the medium between the plates. At room temperature, the dielectric constant of air is approximately 1, that of water is 80, and that of ice is 3–4. Therefore, when the medium between the plates is air, ice, and water, the capacitance reflected by the capacitor will be different. Due to changes in temperature, the dielectric constant of the medium changes, and the capacitance value changes accordingly. In the experiment, the authors used a parallel-plate capacitor to measure the capacitance of air, ice, and water at 11–-20℃, and the results are shown in Figure 1. Analysis of the curve results shows that, since air is a non-polar dielectric, the capacitance value does not change significantly with temperature, remaining basically between 0.2 and 0.4 nF. The capacitance of water and ice is highly sensitive to temperature changes. From 11 to 0°C, the capacitance of water monotonically decreases from 21 to 34 μF. Near zero degrees Celsius, water begins to freeze. Because the physical state of water changes from liquid to solid ice, the dielectric constant changes significantly, causing a one-step change in capacitance. From 0 to -20°C, the capacitance of ice monotonically decreases from 5 to 68 nF. 3. Basic Principle of Ice Thickness Detection Based on the analysis of the capacitance characteristics of air, ice, and water changing with temperature, we use the detection device shown in Figure 2 to achieve continuous automated detection of ice thickness and sub-ice water level. In the detection of river ice or sea ice, three main substances are involved in the three-dimensional environment: air on the ice surface, the ice layer, and sub-ice water. If the area and spacing of the plates of a parallel-plate capacitor are fixed, and the space between the plates is filled with the medium being tested (air, ice, or water), then the capacitance of the parallel-plate capacitor is determined by the dielectric constant εr of the medium being tested. During testing, the parallel plate capacitor and switching circuit shown on the right in Figure 2 are packaged into a cylindrical device and placed vertically at the test point. The plates of the capacitance detection device are filled with subsurface water, ice, and air on the ice surface. Each plate is connected to a chip select switch. Through program control, the microcontroller can control the chip select circuit to sequentially select switches 1 through n. A capacitance/frequency conversion circuit converts the capacitance values ​​of the parallel plates at different vertical heights into corresponding frequency values. To eliminate the influence of changes in relative permittivity on the measurement results, a reference plate is placed at the bottom of the capacitance detection device. The reference plate is located in the water and connected to capacitance/frequency conversion circuit 2. The other detection plates are connected to capacitance/frequency conversion circuit 1 via the chip select circuit. Capacitance/frequency conversion circuits 1 and 2 are identical. The microcontroller receives the frequency values ​​transmitted from the two capacitance/frequency conversion circuits. Because the capacitance characteristics of ice and water differ greatly, the frequency values ​​transmitted from different media will also differ significantly. Using the frequency value of a reference plate as a benchmark, the frequency value of the detection plate is compared with it. When the frequency values ​​of the detection plate and the reference plate are the same or similar, the plate is considered to be in water; otherwise, it is considered to be in ice. When adjacent plates are in ice and water respectively, the lower interface of the ice is determined, and the height of the water under the ice is calculated. Since the dielectric constant of air is basically unaffected by temperature, we can quantify the capacitance range of air separately, convert it into the air frequency range, and use this to determine whether the medium at the plate is air. When adjacent plates are in ice and air respectively, the upper interface of the ice is determined. Based on the upper and lower interfaces of the ice layer, the ice thickness can be calculated. 4. Experiment and Conclusion A parallel plate capacitive sensor was selected, and an ice thickness measurement experiment was conducted according to the above approach, as shown in Figure 3. The parallel plate plates are rectangular copper-clad plates with a length of 1 cm and a width of 3 cm, and the two plates are spaced 1 mm apart. The plates are distributed on both sides of the fixed frame, with one side of the plates sharing a common electrode, and the other side of the plates connected to the chip select circuit. Based on the principle of capacitance detection, the entire electrode does not need to be in direct contact with the medium. The entire sensor is enclosed, with the electrodes and their wires sealed within a fixed frame. The capacitor/frequency conversion circuit employs an RC multivibrator. An MSP430 microcontroller was used for circuit selection, and tests were conducted on two scenarios: a simulated ice-water interface and an ice-air interface. In the lower interface test, the ice frequency was 2.7 MHz and the water frequency was 35 Hz. In the upper interface test, the ice frequency was 2.7 MHz and the air frequency was 4 MHz, as shown in Figures 4-6.