Atmospheric Remote Sensing model refers to the detection methods and technologies that the instruments do not directly contact with the atmosphere and then measure the ingredients, motion states and the meteorological elements' values in a distance.Both weather radars and weather satellites fit into the category of Atmospheric Remote Sensing.
The applications of remote sensing in hydrology and water resource include water resource investigation, watershed planning, watershed area distribution and changes, estimation of runoff, water depth, water temperature, snow cover, soil moisture, ice monitoring, investigation of estuarine coastal zones and offshore topography, marine research, and so on.
The quantitative estimation of forest structure parameters is a main task of remote sensing. The estimation of forest structure parameters at high accuracy should be based on the full understanding of interactions between optical or microwave signals and forest stands which could be achieved by forward modeling of remote sensing data.
Snow cover is an important part on earth surface, 3/4 of the fresh water on earth exits in the form of snow and ice. In winter, 80% of the Eurasia and North America is covered by snow, and the average snow cover area of the hemisphere in January is about 46500000 km2, and 3800000 km2 in August. In high latitude area, snow is the main source of river and underground water.
Soil is one of the most important substance in the Earth system. It’s very important to precisely simulate emissivity of bare soil. Currently, AIEM is an important model to simulate soil emissivity. The three dielectric constant model Mironov, Dboson and Frozen Dielectric model provide the ability to simulate dielectric constant in different conditions.
Crops provides human food, and the output of crops is directly related to food security. The early method is to use the vegetation index method or regression empirical relationship to do the remote sensing monitoring of crops. The advantage of these methods is that it is easy to get. The disadvantage is that the model is not global and the model can not adapt to other regions.