Composition Of NOAA-20 VIIRS Nighttime Light Monthly Data And Its Potential Applications

The acceleration of urbanization process has gradually exposed a number of urban problems.These urbanization challenges promote an urgent need for new theories and technologies to monitor urban dynamics.In the last few decades,the rapid development of remote sensing technologies,especially the state-of-the-art nighttime remote sensing,bring new opportunities and solutions for the study of urban problems.Consequently,there is an growing demand for wide-coverage,long-term,and sustainable nighttime light remote sensing data.And it has become a research hotspot in the field of nighttime light remote sensing.As the successor of S-NPP VIIRS,a new generation of NOAA-20 VIIRS sensor with the capability to acquire nighttime light data was launched in November 2017,which provides more opportunities for obtaining wide-coverage,longterm,and sustainable nighttime light remote sensing data.It is,therefore,important to evaluate the quality of NOAA-20 nighttime light data with reference to the previous nighttime data to support further analysis in future research.However,few studies have been conducted to composite NOAA-20 VIIRS monthly data and evaluate its application in the urban areas.And the research objectives of this study were:(a)to generate monthly NTL composites in China from NOAA-20 VIIRS,(b)to evaluate the consistency in the monthly NTL composites of NOAA-20 and S-NPP at multiple scales,and(c)to Assess its potential applications.The main research contents are as follows:(1)A series of factors that affect the quality of NOAA-20 VIIRS nighttime lights daily data were analyzed.A corresponding denoising process was then designed to eliminate the impacts of cloud,sunlight,moonlight and nonlinear noise at scan margins in NOAA-20 VIIRS nighttime light data.Next,outliers in the time series were further eliminated by employing the modified z-score method.Finally,an average strategy was adopted to composite the daily NOAA-20 nighttime light data,the resultant NOAA-20 VIIRS nighttime light monthly composite dataset in China from April to December,2019 was released at https://doi.org/10.6084/m9.figshare.13625210.v2.(2)The consistency of NOAA-20 VIIRS and S-NPP VIIRS nighttime light monthly data was quantitatively compared and analyzed at the provincial,prefectural and pixel scales.The results show that NOAA-20 VIIRS nighttime light monthly data are highly correlated with S-NPP VIIRS nighttime light monthly data at the provincial and prefectural scale,and the average goodness of fit of the analytical model is 0.99.On the pixel scale,NOAA-20 VIIRS and S-NPP VIIRS nighttime light monthly data also show good consistency in the entire study area,with an average goodness of fit of analytical model reaching 0.86,and the average RMSE being 3.86 n W/cm2/sr.The two datasets are also highly consistent in terms of spatial trends.In general,NOAA-20 VIIRS and S-NPP VIIRS nighttime light monthly data show good consistency at various scales.Potential factors that may affect the consistency between the two datasets,including satellite observation angle,transit time,and the proportion of stable pixels,were also investigated.The results shown that the proportion of stable pixels is the major factor affects the consistency.(3)We used NOAA-20 VIIRS nighttime light monthly composite data to estimate GDP,EPC and extract the urban areas in mainland China.we first used a simple linear regression model to investigate the potential of NOAA-20 nighttime light data for estimating GDP and EPC at multiple scales;we then extracted urban areas from NOAA-20 nighttime light data by using Random Forest model.The results show that NOAA-20 VIIRS nighttime light monthly data achieve good estimation of GDP and power consumption at both provincial and prefectural scale,and the goodness of fit of all estimation models exceeds 0.7.In terms of urban built-up area extraction,NOAA-20 VIIRS nighttime light monthly data also show good performance,and the overall accuracy of built-up area extraction is higher than 85% in all selected cities,with a maximum of 98.3%.All these show that NOAA-20 VIIRS nighttime light monthly composite data have great potential in related applications.To sum up,NOAA-20 VIIRS nighttime light monthly data show good consistency with S-NPP VIIRS nighttime light monthly data and has great application potential.As a result,NOAA-20 VIIRS nighttime light monthly data will play an equally important role in the research field supported by S-NPP VIIRS nighttime light monthly data.Besides,this study can provide a base scientific reference that NOAA-20 VIIRS nighttime light data can be a useful data source for enlightening more applications in the fields of socioeconomic and urban studies.