Study On Seasonal Changes Of VIIRS Nighttime Light Based On Time-series Analysis

Night-time light(NTL)images show a high correlation with human activities and are widely used in many fields.Researchers analyzed the spatial distribution of economic parameters using the spatial information,fitted and predicted the change trend of economic parameters using the time-series data,and assessed the impact of emergencies using the mutations.In all,the spatial information and the trend information in the NTL data has been well recognized and applied,while researches on the seasonal changes in the data are relatively few.Learn more about the contents in the NTL data can help to broaden the application fields,as well as enhance the accuracy of existing applications.This paper focuses on the annual changes of the NTL data,analyzes the spatial distribution of the NTL seasonality,concludes and summarizes the main factors and their effects on it.The main research contents and conclusions are as follows:(1)This paper proposed the method of "seasonal base vector" to quantify the seasonal information of time-series,and compared it with seasonal index method and season-trend decomposition algorithm based on moving average.The analysis shows that the method of seasonal base vector can effectively evaluate the seasonal fluctuation of each period in the time-series,and also can describe the seasonality of the entire time-series.The calculation method is simple,which is suitable for the calculation of large-scale data.(2)This paper calculated the seasonality of the NTL data on the national scale and pixel scale,and analyzed the spatial distribution characteristics.The analysis shows that the seasonality is higher in the north,which is related to the snow cover,the seasonality is relatively low in the low and middle latitudes,while the seasonality in the southern hemisphere is low as well.Some regions with high seasonality in the low and middle latitudes is related to their social culture.(3)The main factors were divided into social factors and natural factors.This paper selected four cities including two Islam holy cities: Macca and Najaf,and two typical cities in Mexico: Cancun and Puebla City,and then studied the seasonality features in the time dimension and space dimension.The analysis shows that the NTL increases significantly during the large-scale cultural festivals,different types of festivals with different celebrations have different change characteristics.The growth is more concentrated during the religious activities while the growth evenly distributed during the Carnival-type festivals.(4)The natural factors on the NTL seasonality were divided into snow cover and vegetation cover.The analysis of the impact of snow cover selected China as the research area,and analyzed the relationship between the NTL seasonality and average NDSI on the provincial scale and pixel scale in four typical cities.The analysis shows that snow cover has a huge impact on the NTL data.On the provincial scale,the NTL seasonality and the average NDSI are at a low level in most provinces,several regions have high values for both and show a linear correlation trend;on the pixel scale,the average NDSI shows no relationship with the seasonality.The seasonality of different land cover types is almost the same,and the increase in snow cover to the NTL data is basically greater than 70%.(5)The study selected deciduous broadleaf type and Nanjing as the study area to analyze the influences of the vegetation cover on the NTL data.The analysis shows that the seasonality of Nanjing urban areas is significantly lower than that of deciduous broadleaf areas.Vegetation coverage has an impact of 12% on the NTL data.