Online first Articles http://www.gyqx.ac.cn EN-US http://www.gyqx.ac.cn/EN/current.shtml http://www.gyqx.ac.cn 5 By using the observation data of the land-atmosphere interaction flux observation tower of a typical secondary evergreen broad-leaved forest in Phoenix Mountain of Zhuhai from October 2015 to July 2016 and the land surface process observation tower of the hilly shrub forest in Zengcheng area of Guangzhou from October 2014 to July 2015, we analyzed the energy flux, Bowen ratio, and carbon flux of the two types of forest underlying surfaces. The results show that the sensible heat at Zengcheng Station dominates in the dry season, and the sensible heat at Zhuhai Station is equivalent to the latent heat. As precipitation increases in the wet season, the latent heat at the two stations dominates; the average diurnal fluctuations in sensible heat and latent heat at Zengcheng Station in the dry season are slightly more gentle than those at Zhuhai Station; Both stations had negative heat during the night in the dry and wet seasons, and inverse humidity occurred at individual moments. The Bowen ratio in the dry and wet seasons fluctuates greatly at night, and the Bowen ratio in the daytime is positive, and the fluctuation is small; the Bowen ratio at Zhuhai Station during the dry season is smaller than that at Zengcheng Station; In the wet season, the daytime Bowen ratio is less than 1. The carbon flux variation range of secondary evergreen broad-leaved forest is larger than that of hilly shrubs; the carbon sequestration capacity of typical secondary evergreen broad-leaved forests is stronger than that of hilly shrubs. The carbon flux at Zengcheng Station in the dry season is smaller than in the wet season; the carbon flux in Zhuhai’s wet season is slightly smaller than that in the dry season, which may be related to the weather conditions at Zhuhai Station and the direction of the flux contribution area.

]]> Radar Quantitative Precipitation Estimation（QPE） can provide continuous, high-resolution and wide-coverage precipitation products which have become critical means of meteorological monitoring. Although the current methods in different regions have made considerable progress, the effect of QPE in complex terrain areas still needs to be strengthened. Aiming at the QPE in the complex terrain area of the northeast slope of the Qinghai-Tibet Plateau, this paper establishes a real-time-dynamic-correction algorithm used geographic terrain factors based on a quasi-stable moving-window based on existing estimation algorithms: fixed empirical Z-I relationship, cloud classification empirical Z-I relationship, and echo classification statistics Z-I relationship. The estimation performance evaluation was conducted using the precipitation events in Linxia during the flood season from 2017 to 2018. The results demonstrate that the estimated precipitation value without the correction algorithm is often smaller than the actual observation value, and the corrected average precipitation value is more consistent with the actual observation value. For small rainfall events, the cumulative distribution function (CDF) obtained by different estimation algorithms deviates greatly from the actual observation. As the precipitation level gradually increases, the CDF of estimated precipitation gradually approaches the actual observation. It is more stable in the estimation performance of the correction algorithm based on the statistics Z-I relationship of the echo classification. The error of precipitation estimation by different algorithms is quite different and the uncorrected algorithm mainly underestimates to all precipitation events. The overall estimation deviation of the correction algorithm both based on Z-I relationship of echo classification statistics and Z-I relationship of fixed experience is relatively small. The uncorrected algorithm has the advantage to estimate the main precipitation area of convective-precipitation-weather events, but it has a poor ability to estimate the precipitation intensity, and the estimated effect is a boost after the correction. In particular, the correction algorithm based on cloud classification experience Z-I relationship has better estimation ability for convective precipitation weather events, especially the estimation ability of short-duration heavy rainfall is significantly better than other algorithms. The uncorrected algorithm has weaker precipitation estimates for stable precipitation weather events, and the corrected estimates get closer to observations.]]>