A Study of The Growth of The Amount of Meteorological Observation Sites over 200 Years

Jin Chai, Wei Jiang, Yuxia Fu, Choujun Zhan, Zhiwen Yu

Abstract: Studies of climate data is an essential topic which has attracted increasing attention from the scientific community. Researchers from disparate academic disciplines investigate climate change and develop techniques for weather forecasting based on reliable climate data. A meteorological observation site is one the major source for providing reliable climate data. Hence, the amount of weather observatories in a country directly reflects the level of research on weather in this country. In this work, we established a dataset of the annual amount of observation sites that operate normally in 181 countries from 1800 to 2017. Furthermore, we collect data on economic, research investment, agriculture investment, and etc, in each country, which may be related to the amount of observation sites. Then, we analyze the correlation coefficient between all kinds of factors (such as GDP and agricultural acreage) and the growth of the amount of observation sites in a country. Experimental results indicate that some of the factors are highly related to the amount of observation sites but some are not. Moreover, we develop neural network models to forecast the amount of sites in a country. Experimental results indicate that the prediction is accurate.