Abstract: Due to lack of detail observations of wire icing processes under natural environment, evolution characteristics of wire icing processes isn't very clear and the modeling is difficult. To solve this problem, under the support of the National Research Funds for Public Welfare, a special equipment for automatic monitoring the mass of ice accretion on wires is successfully trial-manufactured and employed in field observation experiments carried out in Guizhou. During the field observation experiments from January 2011 to March 2013, plenty of detail observations of wire icing process are obtained, including minutely evolution of ice mass and relative meteorological elements, which provide conditions for subsequent research. Using the experiment data, evolution characteristics of wire icing processes and relative meteorological conditions are investigated. Results show that there are five different evolution phases in the whole icing process, including beginning, growth, persistence, fading and dispelling of the ice-coat, and the judgment criteria of meteorological conditions are found out. A numerical model describing wire icing processes caused by rime and glaze is developed and applied to simulate observed wire icing processes.Based on the theoretical model frame for wire icing which is universally accepted, and combined with the judgment criteria of meteorological conditions, a numerical model of wire icing caused by rime and glaze based on the process judgment is presented. The model adopts improved parameterization schemes and algorithms, including respective procedures for rime and glaze icing, and takes the mass loss due to sublimation and surface evaporation into account. With the in-situ meteorological data combined with the judgment criteria of meteorological conditions, the numerical model is validated by applying to simulate observed wire icing processes. Modeling results show it is able to describe correctly the whole icing process, including growth, persistence, fading and dispelling of the ice-coat, with the maximum difference of ice-coat quality between modeling and observation less than 20%. Modeling errors come from two aspects, including calculation errors of the model itself and errors of meteorological data input. The icing model with common meteorological elements input can be used to calculate and hourly variation of mass, thickness and density of the ice-coat, and is of practical value for application. Under the condition of inputting with microphysics observations of real weather processes, the modeling result can be further improved. Furthermore, taking advantage of products from a refined local numerical weather forecast model, this icing model can be applied in predicting the evolution of the ice accretion on the real power transmission lines, and has wide application prospect.