Study of the Variability of Arctic Perennial Ice Based on Satellite Microwave Measurements Using the Neuron Network (1988-2001)

摘要

Arctic perennial ice (PI) is an important component of the climatic system. Therefore, the collection of reliable information on PI distribution and dynamics, as well as its correlation with atmospheric processes and climatic changes is a pressing issue. Data on different aspects of this problem, including those based on satellite-based monitoring, have been discussed in [3, 4, and 9]. Modern satellite-based microwave information technologies serve as an efficient tool for studying different geophysical parameters of sea ice and climate. However, analysis and validation based on the well-known operational algorithms demonstrate a substantial seasonal and regional variability of PI values. For instance, assessments of PI concentration in which the SSM/I data is based on the NASA Team and Bootstrap algorithms demonstrate significant differences. The present paper is devoted to the development of new methods for assessing PI parameters and distribution based on active and passive microwave satellite-based measurements. The relationship of PI dynamics with atmospheric processes and climate is also studied using regression analysis.