Progress for Stable Artificial Lights Distribution Extrvction Accuracy and Estimation of Electric] Power Consumption by Means of DMSP/OLS Nighttime Imagery

The Noise Reduction Filter (NRF) that is developed by the authors is applied to extract artificial nightlight components of a time series DMSP/OLS-VIS dataset. High frequency components from the time series DMSP/OLS-VIS dataset are exhausted and a direct current component is extracted by the NRF that is one of the Fourier analysis techniques. The inference of cloud and other disturbance noise are also removed, and a stable artificial nightlight is extracted by the NRF filtration. The intensity value in high power light areas observed by DMSP/OLS-VIS is saturated because of narrow dynamic range of the sensor gain. A simple model called "Deltaic Model" developed by authors corrected those saturated value. Verification of the accuracy of correction methods above described is carried out by comparison with electric power consumption of the calculated values from the model and statistical ones of each prefecture in Japan. Correlation of the values is satisfactory as shown R2 = 0.725. The results of this work shows the remote sensing method by using the DMSP/OLS-VIS nighttime imagery with the correction methods above described is useful to estimate the electric power consumption through a year of fixed areasp.31-42 : ilus. ; 30 c