Atmospheric visibility refers to the maximum distance at which a person with normal vision can discern and identify a target object against the sky background. It plays a crucial role in aviation, maritime navigation, traffic safety, meteorological observation, and air pollution research. Remote sensing technology has long been a key focus for visibility monitoring. Compared to conventional visibility sensors, lidar has demonstrated significant advantages, including high spatiotemporal resolution, real-time data acquisition, and the capability for single-end detection. These advantages make lidar a promising tool for visibility observation.This study reviews existing visibility measurement methods, highlighting previous research on visibility monitoring using lidar. It further explores a visibility retrieval method based on coherent Doppler wind lidar. A field experiment was conducted in the Shiyan region of Shenzhen from September 9 to October 6, 2019. A coherent Doppler wind lidar was operated alongside the Shenzhen Meteorological Tower for joint observations. The meteorological tower was equipped with nephelometers at three different heights to provide truth visibility measurements.The backscatter intensity measured by the coherent Doppler wind lidar, along with the horizontal visibility and Relative Humidity (RH) data recorded by the nephelometer and its humidity sensor, were utilized to perform nonlinear fitting based on theoretical analysis. To verify the effectiveness of the fitting method, the collected data was divided into training and testing datasets in chronological order. The training data spanned from September 9 to October 1, 2019, while the testing period covered October 2 to October 6, 2019.With training data, a natural logarithm relationship model was established between horizontal visibility and lidar backscatter intensity. This model was then applied to estimate and validate the visibility measurements in the testing dataset. The experimental results indicate that at heights of 50 m and 100 m, both correlation coefficients between the lidar-derived visibility estimates and the observed values exceed0.83 in the testing dataset. These findings demonstrate that coherent Doppler wind lidar offers high accuracy and reliability for visibility measurement, making it a viable alternative to traditional methods. Additionally, to eliminate potential outliers, a 90% confidence interval was applied to filter residual data, further improving the correlation coefficient. This refinement process underscored the robustness of the model and its ability to enhance the precision of lidar-based visibility estimation.Furthermore, based on the classification criteria for fog and haze, the observational data was divided into two categories: one with RH greater than or equal to 80% and the other with RH below 80%. Separate fitting functions were developed for each category, followed by an analysis of their respective characteristics. The classified lidar backscatter intensity data was then used to estimate horizontal visibility for both groups. The results indicate that when RH is below 80%, the correlation coefficient between the estimated and observed visibility exceed 0.8. However, for RH above 80%, the correlation coefficient is slightly lower at 0.79. This reduction in correlation may be attributed to the limited amount of available data under high-humidity conditions, leading to suboptimal model fitting and reduced estimation accuracy. These findings highlight the influence of humidity on visibility estimation and underscore the importance of data availability in improving model performance.Overall, the results suggest that coherent Doppler wind lidar is a feasible and valuable tool for visibility measurement. It provides an effective means of obtaining horizontal visibility data and serves as a reliable remote sensing technique for future visibility monitoring and atmospheric environmental research.