https://tede.unioeste.br/handle/tede/527 2026-09-22T15:45:19Z https://tede.unioeste.br/handle/tede/8652 Title: Distribuição de recursos da cota-parte do ICMS no estado do Paraná com foco na produção agropecuária e na qualidade da educação Autor: Vicente, Daniela Maria Grande Primeiro orientador: Johann, Jerry Adriani Abstract: Municipalities are federative entities whose share of tax revenues is an important element of local public management. Under the Tax on the Circulation of Goods and Services (ICMS), the Brazilian Federal Constitution requires that 25% of state tax revenue be transferred to municipalities. This portion corresponds to the municipal revenue-sharing quota, of which 35% is redistributed according to criteria defined by state legislation, with at least 10% allocated to educational indicators. In the state of Paraná, 8% of the municipal ICMS share is distributed proportionally to each municipality's Gross Value of Agricultural and Livestock Production (VBP) relative to the state total. In this context, formulating public policies to promote greater regional equity poses a challenge, particularly given the diversity and scale of agricultural and livestock production in Paraná. The traditional territorial segmentation into mesoregions and microregions does not necessarily ensure homogeneity in production patterns across municipalities. Therefore, the main objective of this thesis was to develop a methodology to characterize agricultural production across municipalities in the state of Paraná, aiming to form homogeneous groups according to their production patterns. Additionally, the research analyzes the distributive effects of introducing the ICMS Education criterion in the allocation of the municipal revenue-sharing quota, which assigns 10% of the resources based on the Paraná Education Quality Index (IQEP). The thesis is structured into two articles. The first article applies cluster analysis to form homogeneous groups of municipalities, based on the average percentage shares of the Gross Value of Production for 90 agricultural products included in the state agricultural production matrix during the period from 2019 to 2023. The clustering process was conducted in two stages. First, a hierarchical agglomerative method using Ward’s linkage criterion was applied to support the determination of the number of clusters and the estimation of initial centroids. Subsequently, this information was used as input for the K-means algorithm. The analysis identified nine groups of municipalities with similar production profiles. The results show that, on average, the 20 most relevant products account for approximately 90% of the Gross Value of Production within the clusters, indicating a high level of internal homogeneity in production composition. The second article evaluates the distributive effects of incorporating the IQEP into the Municipal Participation Index, the relative importance of each index component, and the short-term fiscal impacts on municipal revenues. The methodology combines descriptive statistics, nonparametric correlation estimates, and cluster analysis to classify municipalities into homogeneous groups according to educational performance, socioeconomic profile, and the size of municipal education systems. The results indicate that the redistribution resulting from the incorporation of the IQEP produced concentrated effects in a limited set of municipalities, without significantly altering, in aggregate terms, the relative participation of most municipalities in the ICMS revenue-sharing system. Furthermore, the findings show that the IQEP calculation methodology is strongly associated with the number of enrollments in municipal school systems. At the same time, educational performance and equity indicators exert only a marginal influence on the final index value. It is concluded that the Paraná ICMS Education model has limited redistributive capacity and a low potential to induce public policies that improve educational quality and equity, largely preserving the existing distributional logic. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Tese 2026-06-26T00:00:00Z https://tede.unioeste.br/handle/tede/8519 Title: AgDataBox-Map-Clean: Computational Module For Data Cleaning Autor: Alves, Fernando de Lima Primeiro orientador: Souza, Eduardo Godoy de Abstract: Os mapas de produtividade estão entre os instrumentos essenciais da agricultura de precisão para avaliar a variabilidade espacial da cultura em uma área agrícola. No entanto, a qualidade dos mapas pode ser reduzida por erros sistemáticos, erros aleatórios e dados atípicos na coleta de dados de campo brutos. Assim, um módulo computacional para limpeza de dados (ADB-Map-Clean) foi desenvolvido e integrado dentro da aplicação AgDataBox-Map (ADB-Map) da plataforma web AgDataBox (ADB; https://adb.md.utfpr.edu.br/ ). ADB é uma plataforma web para integração de dados, software, procedimentos e metodologias para agricultura digital. Este módulo computacional limpa os dados brutos do monitor de colheita (monitor de produção) ao remover os outliers e inliers. Esta tese foi formatada em três artigos: Artigo 1: Foi realizado um estudo sistemático da literatura (SLS) com o objetivo de levantar os métodos utilizados para a remoção de outliers e inliers, bem como o software utilizado na análise exploratória dos dados. O SLS foi conduzido em três etapas: (i) mapeamento sistemático da literatura (SLM), (ii) bola de neve (SB) e (iii) revisão sistemática da literatura (SLR). Foram obtidos 772 trabalhos científicos, os quais foram reduzidos a 86 trabalhos após o processo de seleção. Artigo 2: foi desenvolvido um módulo computacional web (ADB-Map-Clean) para limpeza de dados brutos (usualmente dados do monitor de colheita), capaz de remover valores menores ou iguais a zero, leituras duplicadas, outliers e inliers, integrando-o com o ADB-Map. Em um estudo de caso, uma média de 20% dos dados foram removidos após a limpeza dos dados pelo ADB-Map-Clean, que reduziu significativamente SD e CV, e permitiu a construção de mapas temáticos e zonas de gerenciamento mais precisos. Artigo 3: estudo de caso com dados de três áreas experimentais, onde foram aplicadas diferentes técnicas de remoção de erros associados ao monitor de colheita, utilizando três softwares (i) o módulo computacional web ADB-Map-Clean, (ii) MapFilter e (iii) YieldEditor. Os mapas de produtividade (YMs) construídos com dados processados pelos três softwares de limpeza foram visualmente muito semelhantes, mas sua semelhança com os YMs de dados originais (sem filtragem) era muito pior. Em média, os softwares reduziram o SD em 68% e o CV em 70% após uma redução média de 23% nos dados. Além disso, a produtividade média aumentou 11%. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Tese 2021-09-03T00:00:00Z https://tede.unioeste.br/handle/tede/8518 Title: Qualidade de grãos de milho armazenados em escala comercial: Análise dos efeitos do resfriamento artificial Autor: Mafra, Joyce Bueno Primeiro orientador: Christ, Divair Abstract: Large-scale grain storage in Brazil has been facing challenges imposed by climatic conditions that speed up deterioration and favor pests and fungi proliferation. Artificial cooling of the grain mass emerges as a technological alternative to conventional aeration. Thus, this doctoral dissertation presents an integrated analysis of the effects of artificial cooling on the quality and sanity of 12,000 tons of Zea mays L., (3,000 maize tons per silo in four commercial metallic silos), by comparing artificial cooling with conventional management (aeration and chemical control). The research evaluated the effects on the physical, hygroscopic, and nutritional quality condition of grains (Chapter 1), the population dynamics of pest insects (Chapter 2), the incidence of mycotoxins (Chapter 3), and thermal dynamics based on mathematical modeling (Chapter 4). The results revealed distinct responses dependent on context. For grain quality, cooling showed selective influence dependent on time. For pest insects, efficacy was species-dependent: Sitophilus spp. increased to 28.3 insects/trap (7 times higher than the control with 4.0 insects/trap); Tribolium spp. reached 924.8 insects/trap in the control versus 411.8 in cooling; Cryptolestes spp. presented 242 versus 156 insects/trap; Oryzaephilus spp. reached 577 insects/trap in cooling (8.2 times higher than the control). For mycotoxins, zearalenone showed lower concentrations and greater stability in cooling; while fumonisins did not reduce significantly, but temporal evolution was modulated; and deoxynivalenol increased in both treatments in November (p < 0.001). Mathematical modeling revealed radial thermal heterogeneity, with the silo center up to 7 °C warmer than the periphery from 09/03rd/2023 to 09/08th/2023, consequently, heat pockets were created. Therefore, it is concluded that artificial cooling is a powerful tool, but not universal, and should be integrated into an Integrated Pest Management (IPM) program that considers the spatial heterogeneity of temperature and the ecological characteristics of each target species. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Tese 2026-02-06T00:00:00Z https://tede.unioeste.br/handle/tede/8516 Title: Comparação Multi-Temporal da Temperatura da Superfície Terrestre em Cascavel - PR (2005–2025) Autor: Lucca, Thaynnan Aline Begozzi Liotto Primeiro orientador: Mercante, Erivelto Abstract: The accelerated urban growth of the municipality of Cascavel, Paraná State, Brazil, over recent decades has promoted significant changes in land cover and land use, characterized by the expansion of impervious surfaces and the reduction of vegetation cover. These processes have contributed to increased land surface temperatures and the intensification of urban thermal discomfort. Combined with global climate change, such dynamics increase public health risks and compromise population quality of life, posing a major challenge for local urban and environmental planning. In this context, the objective of this study was to analyze and compare the punctual interannual dynamics of land surface temperature (LST) in the municipality of Cascavel–PR between 2005 and 2025, and its relationship with urban expansion, vegetation cover variation, and geoenvironmental factors. The research was based on the use of geotechnologies and orbital remote sensing techniques, employing Landsat satellite imagery to map land surface temperature (LST), the Normalized Difference Vegetation Index (NDVI), and hypsometry. Statistical and spatial analyses were conducted using the R computing environment. As an original methodological contribution, the study proposed the use of climatic proxies to address the lack of complete historical meteorological time series, through the development of the Human Thermal Comfort Proxy Index (CTH-proxy). This index integrates surface thermal amplitude and monthly precipitation, enabling a retrospective analysis of urban thermal comfort, unprecedented for the municipality. The results indicated a consistent pattern of surface warming in Cascavel–PR, associated with urban expansion and the reduction of vegetation cover. In 2005, thermal stress exhibited a punctual character, whereas in 2015 high variability was observed, influenced by the occurrence of a strong El Niño event. In 2025, greater temporal persistence of thermal stress was identified, indicating a worsening of conditions associated with human thermal comfort. It is concluded that the integration of remote sensing data, geoenvironmental variables, and climatic proxies constitutes a robust approach for analyzing urban thermal comfort, providing relevant technical support for urban planning, urban forestry management, and climate change adaptation strategies in growing medium-sized cities. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Tese 2026-03-05T00:00:00Z