Guia para qualidade : do tratamento ao armazenamento dos dados ambientais

• Ano de defesa: 2021
• Autor(a) principal: Ferreira, Raphael Pires
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Instituto de Física (IF); UFMT CUC - Cuiabá; Programa de Pós-Graduação em Física Ambiental
• Palavras-chave em Português: Qualidade de dados; Variáveis ambientais; Armazém de dados; CNPQ::CIENCIAS EXATAS E DA TERRA::FISICA; Data quality; Environmental variables; Data warehouse
• Link de acesso: http://ri.ufmt.br/handle/1/3450

Obtaining climatological data through automatic sensors has become increasingly trivial, whether due to the simplicity of accessing portals that provide such information, or by purchasing and installing such equipment. However, the massification of these processes presents some challenges, the need to verify and guarantee the quality of this large amount of data, as well using techniques that guarantee the availability of this data so that the research can be carried out with greater precision. In this sense, given the need for reliable bases for obtaining the data, this research aims to systematize methods that make it possible to assess the quality of data or data set of environmental variables, such as air temperature, soil temperature, among others. A notable gain in efficiency with these validations for the researcher is perceived with the automated application of quality routines, as well as with the creation of techniques in an R environment, as shown in the results. The use of these methods showed that improvements in sensors of the National Institute of Meteorology (INMET) should be made, because several data presented quality problems. Also noteworthy is the creation of the open source platform, called “Plataforma de Acesso a Dados Climatológicos” (PADC). This platform aims to guarantee the concepts of structuring databases in a data warehouse (DW) format, presenting the guarantees of an environment dedicated exclusively to data delivery, where the insertion of new information is strictly controlled. Through this structure, traditional guarantees of relational bases are obtained, as the concept of ease of data table structures and traditional backup techniques can be used. Thus, ensuring more time for the research product to be carried out more effectively. In addition, the dynamic panel of open data enables greater visibility, transparency and customization for faster and improved analysis. In this way, the quality of the data is preserved, not only in techniques that work in the context of the data, but also in the form of its long-term storage, ensuring that this information will not be changed. Several benefits are highlighted when using the platform, including time savings, when it allows dedication to the work methodology and not necessarily in the preparation of data. It is also possible to see results with less bias and therefore more reliable.