两种夜光数据测算社会经济指标能力对比

Comparison of two types of nighttime light data in capacities of estimating socio-economic indicators

为了评估珞珈一号01星夜间灯光数据对社会经济指标的拟合能力,该文将其和美国NPP/VIIRS夜间灯光数据作对比,在全国范围从省级和地市级尺度建立了夜间灯光数据与经济、人口和资源能源3类社会经济指标的线性回归模型。结果表明:两种夜间灯光数据均能较好地拟合经济、人口和资源能源参量,对地区生产总值的线性拟合效果最好,资源能源指标次之;在省级尺度,Luojia 1-01数据与经济和人口指标拟合的R^(2)优于NPP/VIIRS数据,并且各项指标的相对误差表现更优;在地市级尺度,Luojia 1-01数据也达到了目前应用最为广泛的NPP/VIIRS数据的水平。证明了Luojia 1-01数据可靠性和应用潜力。Luojia 1-01数据凭借高分辨率的优势,未来会有更广阔的应用空间。

Aiming to evaluate the capacity of China’s Luojia 1-01 Nighttime light(NTL)data to model socio-economic indicators,this paper established linear regression models of NTL data and six socio-economic indicators nationwide from the provincial and prefecture scales,taking US NPP/VIIRS data for comparison.The indicators belong to the three categories of economy,population,and resource&energy.The results show that:1)Both NTL data perform well in modeling socio-economic parameters.The linear fitting effect with Gross Regional Production is the best,followed by that of resource&energy indicators;2)At the provincial scale,The R^(2)of the regression model between Luojia 1-01 data and economy and population indicators is higher than that of NPP/VIIRS data.Luojia 1-01 data also has a better performance in terms of relative errors.At the prefecture scale,Luojia 1-01 data reaches the level of NPP/VIIRS data,which is most widely applied all over the world.It confirms the reliability and application potential of the Luojia 1-01 data.With the advantage of high resolution,Luojia 1-01 data will have a broader application space in the future.