The reference evapotranspiration (ETo) has long been used as a climate parameter for many studies in climatology and hydrology. However, many regions suffer from shortage of both meteorological monitoring stations and historical information on ETo. Thus, the objective of this study was to develop a daily gridded reference evapotranspiration data set for Brazil that matches the period and grid cells of the Global Precipitation Measurement (GPM) data. ETo was calculated using data from 849 weather stations over the period from 1 June 2000 to 31 December 2018. The features used to model ETo were the GPM daily data set, WorldClim averages monthly, and two engineered features. Among the machine learning algorithms assessed, the Cubist presented the best performance‐computation cost trade‐off in a subset of the entire data and, therefore, was selected to model ETo daily. The developed data set presented root mean square error of 0.65 mm day⁻¹, or 16% lower than previous ETo data set developed for Brazil using interpolation techniques. The GPM and engineered features showed higher importance for the models trained during the wet season, while the WorldClim maximum temperature averages monthly were more important during the dry and cold season. The new gridded reference evapotranspiration data set for Brazil (ETo‐Brazil) was made freely available to the community.

Residue decomposition from cattle dung is crucial in the nutrient cycling process in Integrated Crop–Livestock Systems (ICLS). It also involves the impact of the presence of trees exerted on excreta distribution, as well as nutrient cycling. The objectives of this research included (i) mapping the distribution of cattle dung in two ICLS, i.e., with and without trees, CLT and CL, respectively, and (ii) quantification of dry matter decomposition and nutrient release (nitrogen—N, phosphorus—P, potassium—K, and sulphur—S) from cattle dung in both systems. The cattle dung excluded boxes were set out from July 2018 to October 2018 (pasture phase), and retrieved after 1, 7, 14, 21, 28, 56 and 84 days (during the grazing period). The initial concentrations of N (~19 g kg⁻¹), P (~9 g kg⁻¹), K (~16 g kg⁻¹), and S (~8 g kg⁻¹) in the cattle dung showed no differences. The total N, P, K and S released from the cattle dung residues were less in the CLT system (2.2 kg ha⁻¹ of N; 0.7 kg ha⁻¹ of P; 2.2 kg ha⁻¹ of K and 0.6 kg ha⁻¹ of S), compared to the CL (4.2 kg ha⁻¹ of N; 1.4 kg ha⁻¹ of P; 3.6 kg ha⁻¹ of K and 1.1 kg ha⁻¹ of S). Lesser quantities of cattle dung were observed in the CLT (1810) compared to the CL (2652), caused by the lower stocking rate, on average, in this system (721 in the CL vs. 393 kg ha⁻¹ in the CLT) because of the reduced amount of pasture in the CLT systems (−41%), probably due to light reduction (−42%). The density of the excreta was determined using the Thiessen polygon area. The CL system revealed a higher concentration of faeces at locations near the water points, gate and fences. The CLT affects the spatial distribution of the dung, causing uniformity. Therefore, these results strengthen the need to understand the nutrient release patterns from cattle dung to progress fertilisation management.

One of the obstacles in monitoring agricultural crops is the difficulty in understanding and mapping rapid changes of these crops. With the purpose of addressing this issue, this study aimed to model and fuse the Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) using Landsat-like images to achieve daily high spatial resolution NDVI. The study was performed for the period of 2017 on a commercial farm of irrigated maize-soybean rotation in the western region of the state of Bahia, Brazil. To achieve the objective, the following procedures were performed: (i) Landsat-like images were upscaled to match the Landsat-8 spatial resolution (30 m); (ii) the reflectance of Landsat-like images was intercalibrated using the Landsat-8 as a reference; (iii) Landsat-like reflectance images were upscaled to match the MODIS sensor spatial resolution (250 m); (iv) regression models were trained daily to model MODIS NDVI using the upscaled Landsat-like reflectance images (250 m) of the closest day as the input; and (v) the intercalibrated version of the Landsat-like images (30 m) used in the previous step was used as the input for the trained model, resulting in a downscaled MODIS NDVI (30 m). To determine the best fitting model, we used the following statistical metrics: coefficient of determination (r²), root mean square error (RMSE), Nash–Sutcliffe efficiency index (NSE), mean bias error (MBE), and mean absolute error (MAE). Among the assessed regression models, the Cubist algorithm was sensitive to changes in agriculture and performed best in modeling of the Landsat-like MODIS NDVI. The results obtained in the present research are promising and can enable the monitoring of dynamic phenomena with images available free of charge, changing the way in which decisions are made using satellite images.