A long-term in situ observation was carried out in the Ice-valley at Milyang in order to explain the factors and processes associated with the summertime ice formation. The variation of temperature inside Ice-valley in relation with ice formation in summer time was found to depend on precipitation rate in spring and cold air sinking in autumn and winter. The rate of temperature rising tends to correspond to sensible heat release depending on the precipitation amount at the freezing location. The reason of the cold air accumulation in a talus in the Ice-valley is the cold air sinking over the surface of talus due to the occurrence of outside clod air mass and the accumulated cold air from autumn to spring flow outside at the bottom of talus. The out-flowing cold air can result in the ice formation in the hot summer.

Observations were conducted in Uiseong Binggye-ri Ice Valley of Korea, where a low temperature talus is located. The talus temperature at Ice Hole near the lower end of the talus remained at 0°C until summer or autumn and averaged -0.3°C over 2 years. The talus temperature at Warm Hole was much higher than the external air temperature during autumn and winter. The outflow of warm air at Warm Hole begins in the late summer of fall. The direction of airflow through the talus surface was determined by the relationship between talus temperature and external air temperature. Annual variations in talus temperature and air exchange between the talus and the external environment can be divided into four periods. It was found that the airflow directions at Warm Hole and Ice Hole were not necessarily opposites. The outflows from the talus at Ice Hole and Warm Hole can occur simultaneously.

The meteorological elements were measured to investigate cause of summertime ice formation at Unchiri, Gangwon Province. The cause of freezing at valley was conformed as adiabatic expansion theory, latent heat of evaporation, natural convective theory, cold air remain theory, and convective freezing theory according to former study. However nither theory produced a satisfactory explanation. This studying area is not valley but ridge, and underground water surface exists at below than freezing height. wintertime temperature drop and summertime cold air spouting were explain as natural convective theory, generation of water drop on the rock was explained as cooling theory by air expansion, and ice formation on the rock was explained as adiabatic expansion theory. In conclusion, formation of ice valley at Unchiri was formed by natural convective theory, adiabatic expansion theory, and latent heat of evaporation successively.