Spatial Interpolation and Temperature Information Visualization
Spatial interpolation on temperature field has gained increased interest in recent years. In this paper we investigate the interpolation accuracy of three frequently used methods (i.e. Inverse-Distance Weighting, Thin Plate Spline and Ordinary Kriging) on
- PDF / 496,089 Bytes
- 7 Pages / 439.363 x 666.131 pts Page_size
- 64 Downloads / 210 Views
Abstract. Spatial interpolation on temperature field has gained increased interest in recent years. In this paper we investigate the interpolation accuracy of three frequently used methods (i.e. Inverse-Distance Weighting, Thin Plate Spline and Ordinary Kriging) on the United States temperature data using crossvalidation technique. Our results indicate that Ordinary Kriging method has higher interpolation accuracy than other two methods. Finally, an elevation optimization is added to Ordinary Kriging and we can find that it increases interpolation accuracy even further. Based on the interpolated data, this paper also gives a two-dimensional visualization of temperature data and temperature difference data on United States. Keywords: Spatial interpolation, Ordinary Kriging, algorithm improvement, data visualization.
1
Introduction
In three-dimensional space, the feature data of research object can be obtained through a variety of means and methods, but the data obtained is very limited and the space distributions of data collection points are discrete forms. Due to these characteristics, spatial interpolation becomes a research topic to researchers. Many methods are used to resolve the important issue of spatial interpolation in recent years. How to utilize the limited observational data to get the reasonable spatial distribution of data has become a challenge. This paper primarily researched to spatial interpolation of temperature data. We will insert the monthly maximum and minimum temperature data into the United States. It is generally recognized that the main factors that affect temperature are latitude, longitude and elevation. So this paper is mainly on these two points to optimize the algorithm. Firstly, this paper uses Inverse-Distance Weighting [1][2], Thin Plate Spline [3] and Ordinary Kriging [4][5] method for data interpolation and choose the one that has more highly interpolation accuracy, and then optimizes it at the elevation direction. Finally, the interpolation data are obtained. However, the data obtained are very huge and it is very difficult to analysis directly. So this paper James J. (Jong Hyuk) Park et al. (eds.), Future Information Technology, Lecture Notes in Electrical Engineering 309, DOI: 10.1007/978-3-642-55038-6_28, © Springer-Verlag Berlin Heidelberg 2014
181
182
L. Wang, M. Che, and J. Li
proposes a visualization [6][7] method to display the temperature data. By this way, the temperature data and temperature different data can be displayed at the same time.
2
Data
There are nearly 80 cities observation data within the monthly minimum and maximum temperature data of the United States in 2012. This paper takes the data of December as an example to study. The geographical scope of interpolation is in 25°50°E and 65°-125°N. The data visualization part also uses the United States State map which is provided by local ArcGIS Server.
3
Selection of Interpolation Methods
Spatial interpolation is a mathematical method based on known data of study area to estimate or predict unkno
Data Loading...
