Methods for Scaling Gridded Data

• Published in: Scale in Spatial Information and Analysis pp. 143 - 160
• Main Authors: Zhang, Jingxiong, Atkinson, Peter, Goodchild, Michael F.
• Format: Book Chapter
• Language: English
• Published: United States CRC Press 2014; Taylor & Francis Group
• Subjects: ENVIRONMENTAL ENGINEERING & TECHNOLOGY; Geographical information systems (GIS) & remote sensing; Human geography; GAC Data; GIFOV; Image Deconvolution; Coarse Resolution Data; Moderate Spatial Resolution Data; Blind Deconvolution; Super-resolution Techniques; Finer Resolution Data; Super-resolution Mapping; Wiener Filter; Geostatistical Methods; Linear Degradation Model; Finer Resolution Image; LSF; Gridded Data; Spatial Structural Information; AR Model Coefficient; Coarse Resolution Pixels; MODIS; Area Class Mapping; Landsat MSS; Coarser Spatial Resolution; Convolution Kernel; Sensor Point Spread Functions; Coarser Resolution Images
• ISBN: 9781439829370; 1439829373
• DOI: 10.1201/b16751-11

Scaling or change of scale is important for information transfer across scales, wherever analysis, modeling, or applications are performed on scales different from those of the given datasets, models, or problem domains (Gelfand et al. 2001; Hufkens et al. 2008). As in previous chapters, we focus on methods for (spatial) scaling of spatial data. Chapter 6 described geostatistical approaches to change of scale: upscaling can be performed through block kriging, whereas downscaling can be implemented using block-to-point (or larger support-to-smaller support) kriging, respectively. For downscaling, geostatistical inverse modeling approaches may be usefully explored and provide a motivation for this chapter, as both are explicitly oriented to gridded data, such as images and raster data. Below, we review briefly the geostatistical methods for scaling before moving to the cases of gridded lattice data, highlighting the complementarity of methods covered in Chapters 6 and 7.