TY - JOUR
T1 - Principles and methods of scaling geospatial Earth science data
AU - Ge, Yong
AU - Jin, Yan
AU - Stein, A.
AU - Chen, Yuehong
AU - Wang, Jianghao
AU - Wang, Jinfeng
AU - Cheng, Qiuming
AU - Bai, Hexiang
AU - Liu, Mengxiao
AU - Atkinson, Peter M.
PY - 2019/10/1
Y1 - 2019/10/1
N2 - The properties of geographical phenomena vary with changes in the scale of measurement. The information observed at one scale often cannot be directly used as information at another scale. Scaling addresses these changes in properties in relation to the scale of measurement, and plays an important role in Earth sciences by providing information at the scale of interest, which may be required for a range of applications, and may be useful for inferring geographical patterns and processes. This paper presents a review of geospatial scaling methods for Earth science data. Based on spatial properties, we propose a methodological framework for scaling addressing upscaling, downscaling and side-scaling. This framework combines scale-independent and scale-dependent properties of geographical variables. It allows treatment of the varying spatial heterogeneity of geographical phenomena, combines spatial autocorrelation and heterogeneity, addresses scale-independent and scale-dependent factors, explores changes in information, incorporates geospatial Earth surface processes and uncertainties, and identifies the optimal scale(s) of models. This study shows that the classification of scaling methods according to various heterogeneities has great potential utility as an underpinning conceptual basis for advances in many Earth science research domains.
AB - The properties of geographical phenomena vary with changes in the scale of measurement. The information observed at one scale often cannot be directly used as information at another scale. Scaling addresses these changes in properties in relation to the scale of measurement, and plays an important role in Earth sciences by providing information at the scale of interest, which may be required for a range of applications, and may be useful for inferring geographical patterns and processes. This paper presents a review of geospatial scaling methods for Earth science data. Based on spatial properties, we propose a methodological framework for scaling addressing upscaling, downscaling and side-scaling. This framework combines scale-independent and scale-dependent properties of geographical variables. It allows treatment of the varying spatial heterogeneity of geographical phenomena, combines spatial autocorrelation and heterogeneity, addresses scale-independent and scale-dependent factors, explores changes in information, incorporates geospatial Earth surface processes and uncertainties, and identifies the optimal scale(s) of models. This study shows that the classification of scaling methods according to various heterogeneities has great potential utility as an underpinning conceptual basis for advances in many Earth science research domains.
KW - ITC-ISI-JOURNAL-ARTICLE
KW - 22/4 OA procedure
U2 - 10.1016/j.earscirev.2019.102897
DO - 10.1016/j.earscirev.2019.102897
M3 - Article
SN - 0012-8252
VL - 197
SP - 1
EP - 17
JO - Earth-science reviews
JF - Earth-science reviews
M1 - 102897
ER -