Mapping within­field variability of soybean evapotranspiration and crop coefficient using the Earth Engine Evaporation Flux (EEFlux) application.

Accurate information about the spatiotemporal variability of actual crop evapotranspiration (ETa), crop coefficient (Kc) and water productivity (WP) is crucial for water efficient management in the agriculture. The Earth Engine Evapotranspiration Flux (EEFlux) application has become a popular approach for providing spatiotemporal information on ETa and Kc worldwide. The aim of this study was to quantify the variability of water consumption (ETa) and the Kc for an irrigated commercial planting of soybeans based on the EEFlux application in the western region of the state of Bahia, Brazil. The water productivity (WP) for the fields was also obtained. Six cloud-free images from Landsat 7 and 8 satellites, acquired during the 2016/17 soybean growing season were used and processed on the EEFlux platform. The ETa from EEFlux was compared to that of the modified FAO (MFAO) approach using the following statistical metrics: Willmot's index of agreement (d-index), root mean square error (RMSE), mean absolute error (MAE) and mean bias error (MBE). The Kc from EEFlux was compared to the Kc used in the soybean field (Kc FAO-based) and to the Kc values obtained in different scientific studies using the d-index. A similar procedure was performed for WP. Our results reveal that EEFlux is able to provide accurate information about the variability of ETa and the Kc of soybean fields. The comparison between ETa EEFlux and ETa MFAO showed good agreement based on the d-index, with values of 0.85, 0.83 and 0.89 for central pivots 1, 2 and 3, respectively. However, EEFlux tends to slightly underestimate ETa. The Kc EEFlux showed good accordance with the Kc values considered in this study, except in phase II, where a larger difference was observed; the average WP of the three fields (1.14 kg m-3) was higher than that in the majority of the previous studies, which is a strong indicator of the efficient use of water in the studied soybean fields. The study showed that EEFlux, an innovative and free tool for access spatiotemporal variability of ETa and Kc at global scale is very efficient to estimate the ETa and Kc on different growth stages of soybean crop.

GIS applied to location of fires detection towers in domain area of tropical forest.

In most countries, the loss of biodiversity caused by the fires is worrying. In this sense, the fires detection towers are crucial for rapid identification of fire outbreaks and can also be used in environmental inspection, biodiversity monitoring, telecommunications mechanisms, telemetry and others. Currently the methodologies for allocating fire detection towers over large areas are numerous, complex and non-standardized by government supervisory agencies. Therefore, this study proposes and evaluates different methodologies to best location of points to install fire detection towers considering the topography, risk areas, conservation units and heat spots. Were used Geographic Information Systems (GIS) techniques and unaligned stratified systematic sampling for implementing and evaluating 9 methods for allocating fire detection towers. Among the methods evaluated, the C3 method was chosen, represented by 140 fire detection towers, with coverage of: a) 67% of the study area, b) 73.97% of the areas with high risk, c) 70.41% of the areas with very high risk, d) 70.42% of the conservation units and e) 84.95% of the heat spots in 2014. The proposed methodology can be adapted to areas of other countries.

Applying GIS to develop a model for forest fire risk: A case study in Espírito Santo, Brazil.

A forest fire risk map is a basic element for planning and protecting forested areas. The main goal of this study was to develop a statistical model for preparing a forest fire risk map using GIS. Such model is based on assigning weights to nine variables divided into two classes: physical factors of the site (terrain slope, land-use/occupation, proximity to roads, terrain orientation, and altitude) and climatic factors (precipitation, temperature, water deficit, and evapotranspiration). In regions where the climate is different from the conditions of this study, the model will require an adjustment of the variables weights according to the local climate. The study area, Espírito Santo State, exhibited approximately 3.81% low risk, 21.18% moderate risk, 30.10% high risk, 41.50% very high risk, and 3.40% extreme risk of forest fire. The areas classified as high risk, very high and extreme, contemplated a total of 78.92% of heat spots.

GIS applied to agriclimatological zoning and agrotoxin residue monitoring in tomatoes: A case study in Espírito Santo state, Brazil.

Searches related to global warming have provided important insights into the response of terrestrial ecosystems, but few have examined the impacts on agricultural crops, particularly those associated with the monitoring of agrotoxin residues. In this context, the agriclimatological zoning is an important tool in the planning and consolidation of crops and should be considered in any initiative that involves such planning. This tool is particularly important in the analysis of agrotoxin residues and may be applied by the Program Analysis of Agrotoxin Residues in Food (PARA) created by the National Health Vigilance Agency of Brazil (ANVISA), which enables greater food security and contributes to the improvement of human health. The aim of this study was to elaborate the current and future agriclimatological zoning for the tomato crop, relating it with the monitoring of samples collected by PARA in Espírito Santo State, Brazil. The results indicate that a temperature increase of 5 °C creates a decrease in apt areas from 37.3% to 4.3%, for a total reduction of 33 percentage points (-88.5%). It is noted that of the 41 producing municipalities, only 26 have apt areas greater than 50%, highlighting the municipalities with apt areas greater than 90%, represented by Mantenópolis (100%), Guaçuí (98.5%), São José do Calçado (97.8%), Irupi (94.4%), Santa Teresa (92.3%), and Marechal Floriano (91.4%). The veracity of agriclimatological zoning is proved by a Kendall rank correlation coefficient of 0.876, indicating that the distribution of the variables of apt areas and productivity are similar at the significance level of 0.05 with a confidence interval 95%. After validation of the agriclimatological zoning for the tomato crop, it is recommended that the PARA should monitor 36 municipalities rather than the current 18, representing an increase of 100%. The methodology can be adjusted to agricultural crops of other countries.