ShipIR 입력을 위한 기상데이터 분석

Abstract

A statistical analysis of the US Navy Marine Climate Atlas of the World (USNMCAW) was presented at the 6th ITBM&S Workshop, proposing to pre-select three statistically significant data input conditions (5%, 50%, 95%) that would produce a maximum, average, and minimum platform signature. The methodology has been shown to fail for two important reasons: the effects of air temperature and humidity on direct and indirect solar irradiation were not considered, and the source data did not contain any specific information about the correlation between the various climatic variables (5x). This paper presents a new methodology to analyse the climatic data for input to ShipIR. Historical hourly data from a weather buoy (in the region of interest) are used to come up with a small number of scenarios (N=100) that have the same statistics (CDF, PDF) as the original data set (N=49072). A new data selection algorithm is described which uses a coarse bin (1/3) to subdivide the variable space (35=243 bins), and an ordered approach to ranking the bins and selecting individual points so that a maximum coverage of each variable is achieved (1/100). maximum, average, and minimum platform signature. The methodology has been shown to fail for two important reasons: the effects of air temperature and humidity on direct and indirect solar irradiation were not considered, and the source data did not contain any specific information about the correlation between the various climatic variables (5x). This paper presents a new methodology to analyse the climatic data for input to ShipIR. Historical hourly data from a weather buoy (in the region of interest) are used to come up with a small number of scenarios (N=100) that have the same statistics (CDF, PDF) as the original data set (N=49072). A new data selection algorithm is described which uses a coarse bin (1/3) to subdivide the variable space (35=243 bins), and an ordered approach to ranking the bins and selecting indi