Contenido
El análisis espacial se enfrenta a muchos problemas fundamentales en la definición de sus objetos de estudio, en la construcción de las operaciones analíticas que se utilizará, en el uso de ordenadores para el análisis, en las limitaciones y particularidades de los análisis que se conocen, y en la presentación resultados de analíticas. Muchos de estos temas son sujetos activos de la investigación moderna.
Existen errores comunes que suelen surgir en el análisis espacial, algunos debido a la matemática del espacio, algunos debido a las formas particulares de los datos se presentan espacialmente, algunos debido a las herramientas que están disponibles. Los datos del censo, debido a que protege la privacidad individual agregando datos en unidades locales, plantea una serie de cuestiones estadísticas. La naturaleza fractal de la costa hace que las mediciones precisas de su longitud sean difíciles, si no imposibles. Un software que ajusta las líneas rectas a la curva de un litoral, puede calcular fácilmente las longitudes de las líneas que define. Sin embargo, estas líneas rectas pueden no tener un significado inherente en el mundo real, como se demostró para la costa de Gran Bretaña.
Estos problemas representan un desafío en el análisis espacial debido al poder de los mapas como medios de presentación. Cuando los resultados se presentan como mapas, la presentación combina datos espaciales que son generalmente precisos con resultados analíticos que pueden ser inexactos, dando lugar a una impresión de que los resultados analíticos son más precisos de lo que los datos indican.[1].
Spatial characterization
Defining the spatial presence of an entity limits the possible analysis that can be applied to that entity and influences the final conclusions that can be reached. While this property is fundamentally true of all analyses, it is particularly important in spatial analysis because the tools for defining and studying entities favor specific characterizations of the entities being studied. Statistical techniques favor the spatial definition of objects as points because there are very few statistical techniques that operate directly on line, area, or volume elements. Computational tools favor the spatial definition of objects as homogeneous and separate elements due to the limited number of database elements and computational structures available, and the ease with which these primitive structures can be created.
Spatial dependence or auto-correlation
Spatial dependence is the co-variation of properties within geographic space: features at proximal locations appear to be correlated, positively or negatively. Spatial dependence leads to the problem of spatial autocorrelation in statistics, since, like temporal autocorrelation, this violates standard statistical techniques that assume independence between observations. For example, regression analyzes that do not compensate for spatial dependence may have unstable parameter estimates and produce unreliable tests of significance. Spatial regression models capture these relationships and do not suffer from these weaknesses. It is also appropriate to consider spatial dependence as a source of information rather than something to be corrected[2].
Locational effects also manifest as spatial heterogeneity, or the apparent variation in a process with respect to location in geographic space. Unless a space is uniform and unlimited, each place will have some degree of uniqueness in relation to the other places. This affects spatial dependency relationships and therefore spatial processing. Spatial heterogeneity means that global parameters estimated for the entire system may not adequately describe the process at a given location.
Spatial sampling
Spatial sampling allows determining a limited number of locations in geographic space in order to faithfully measure phenomena that are subject to dependence and heterogeneity. Dependence suggests that, given one location, the value of a different location can be predicted and observations at both locations are not necessary. Heterogeneity suggests that this relationship may change across space and therefore it is not possible to rely on a certain degree of dependence beyond a region that may be small. Basic spatial sampling schemes include the following modalities: random, grouped and systematic. These basic schemes can be applied at multiple levels according to a certain spatial hierarchy (e.g., urban area, city, neighborhood). It is also possible to exploit ancillary data, for example, using property values as a guide in a spatial sampling scheme to measure educational attainment and income. Spatial models such as autocorrelation statistics, regression, and interpolation can also determine the sampling design.
• - Abler, R., Adams, J. and Gould, P. (1972). Spatial organization the geographer's view of the world. London: Prentice-Hall.[[1]](https://www.e-education.psu.edu/geog571/sites/www.e-education.psu.edu.geog571/files/documents/Spat ial%20Organization_The%20Geographer%27s%20View%20of%20the%20World_Abler%20Adams%20and%20Gould_1971_SCANNED%20COPY.pdf).
• - Anselin, L. (2010). Local Indicators of Spatial Association-LISA. Geographical Analysis, 27(2), pp.93-115 [2].
• - Banerjee, S., Carlin, B. P., & Gelfand, A. E. (2004). Hierarchical modeling and analysis for spatial data. Boca Raton: Chapman & Hall/CRC.
• - Benenson, I., & Torrens, P. M. (2005). Geosimulation: Automata-based modeling of urban phenomena. Chichester: Wiley [3].
• - Buzai, G.D. & Montes Galbán, E. (2021) Spatial Statistics: Fundamentals and application with Geographic Information Systems. Buenos Aires: National University of Luján, Geographic Research Institute [4].
• - Fotheringham, A. S., Brunsdon, C., & Charlton, M. (2007). Quantitative geography: Perspectives on spatial data analysis. Los Angeles: SAGE [5].
• - Hypergeo. Spatial analysis [6].