Enfoque metodológico para la construcción de una Geobase como apoyo a la investigación en agricultura y recursos naturales

ResumenLa realización de operaciones con sistemas de información geográfica (SIG) y sistemas de análisis de imágenes de satélite, producen resultados que se insertan y manejan en bases de datos geoespaciales; estos repositorios se convierten en el centro de aplicaciones, en este caso, relacionadas con la investigación en agricultura y recursos naturales. Cada sistema cuenta con un modelo de datos propio y funciones específicas de manejo de sus bases de datos. Al no existir aún un modelo de datos ni funciones estándar para esas bases de datos comúnmente aceptados, el compartir y reutilizar resultados de diversas aplicaciones es un proceso tedioso y que requiere intervención manual del usuario. Esto se dificulta porque numerosos productores de información geográfica no proporcionan suficientes datos descriptivos y concisos (metadatos) acerca de los resultados de sus trabajos. Este artículo aborda una metodología para la construcción de la base de datos geoespacial “GeoBase-L9”, que facilita la construcción del repositorio y el intercambio de productos geográficos entre usuarios. La metodología se basa en un esquema de metadatos propio que permite implementar un sistema de visualización de los insumos y de los productos de investigación vía web.AbstractGeospatial databases are a central part in applications and projects within the context of research in agriculture and natural resources. Performing operations with geographic information systems and with software for the enhancement, display or analysis of geographic data and information contained in digital maps, satellite images and aerial photographs generate results which are inserted within geospatial databases. These repositories are the core of geo-based applications, in particular, those related with a variety of research topics. Each system manages geospatial databases according to its own data model, and performs specific database management functions according to the procedures requested by users. A data model is an essential tool either for software developers or for users. A data model assists in the comprehension, at different levels of depth and detail, of the contents, organization and capabilities of databases, as well as to the comprehension of software restrictions and functionalities required to access the data. Current commercial informatics technology is designed for relational which are the most common types in use; in general this technology is not oriented to the needs of geospatial databases.Relational databases are the most common data repositories in managerial or business applications. In these databases the DBMS or data base management system is based on the relational model, a formal prescription for the organization and structure of data introduced in the 1970s by E. F. Codd. The relational model and the associated Structured Query Language (SQL) are recognized standards for relational databases. Thus, relational DBMS all are based on a common data model which means that at the logical and at the user levels the data are seen through a same lens, which facilitates data exchange among systems and users. Object databases are oriented towards applications where data are inherently more complex than relational tables, such as the data in digital maps. There is no standard object data model, but basic object-oriented concepts for structuring and relating objects such as aggregation, semantic links and the inheritance of properties are commonplace.Following recommendations by the Open Geospatial Consortium, some proprietary database management systems have included in recent years basic constructs to manage geospatial information. Anyway most database management systems, relational or otherwise, do not make public how they manage the internal storage of data items nor how the manipulation functionalities are implemented, since these are trade secrets.The lack of standards for geospatial databases makes the exchange, sharing and reutilization of results among diverse applications and among different users a very tedious process. Normally manual user intervention is required and the task is made difficult and prone to errors because many geographic information producers do not provide sufficient concise descriptive data (metadata) about the results of their work. Most systems offer export/import functions to move geospatial data to/from other systems, but the verification of compatibility for appropriate use of the data is left to the user. She is made responsible for correction or modification of aspects which are significant regarding the applicability of processes or functions. These aspects include file format compatibilities or cartographic projections which when misunderstood lead to repetitive and accumulative errors, with a negative impact on the quality of end products.Multidisciplinary and interdisciplinary research in Geomatics is present in many organizations and institutions worldwide. The objective of the research and development project reported here was to analyze ways to gather in a unique repository geospatial information products to be shared among the group of people working within the research line “Geomatics Applied to the Study and Management of Natural Resources and Agrosystems” of the Colegio de Postgraduados. In the medium term the repository should be able to support decision making using products derived from the research of this group.This article refers to the methodology to build the geospatial database “GeoBase-L9”, which facilitates the construction of the repository and the exchange of geographical products among users. We present the part of the methodology that is based on our own metadata schema which adheres to Mexican national standards which in turn follow international standards. The methodology has allowed the implementation of a pilot version to visualize via web the items that have been input to the GeoBase-L9 and also check the research products offered to the user