Dataset from a System for Collection, Storage and Analysis of Evapotranspiration Data Using Sensors and Artificial Intelligence for Irrigation Optimization and Sustainability in Agricultural Production in Arid Areas

Description

Background: The growing demand for technological advances aimed at more effective use of water resources is crucial to avoid shortages of drinking water and ensure the sustainability of agricultural production in arid regions. Objectives: The data are the result of research and development, the study of which proposes the development of an evapotranspiration monitoring system using sensors and artificial intelligence, with the aim of optimizing irrigation and contributing to sustainability, reducing the risk of environmental degradation. Methods: The research focused on the development of a system that would enable the collection, storage and transmission of data on evapotranspiration on rural properties. To this end, objectives were established that included the construction of a hardware infrastructure dedicated to data communication and the development of a software system equipped with the specific DATASOIS protocol. The experimental methodology used involved the use of advanced research techniques, with low-cost sensors and an innovative virtual platform. Results: Real-time analysis of water consumption allowed the creation of an accessible system that improves water management and prevents soil degradation in irrigated fruit growing areas. This solution aims to promote the sustainable and efficient management of water and soil, combating environmental degradation caused by inadequate irrigation. Financing: The project was made possible with financial support from CAPES and CNPq, contributing to more sustainable and efficient agriculture. Keywords: Evapotranspiration. Data communication. Sensors. Technology. Moisture. Subject The constant data refer to the area of: Agricultural Engineering Specific Area Agribusiness and Agricultural Engineering Data acquisition, processing and analysis system with a low-cost sensor network for the efficient use of water in agriculture. CNPq Process: 434987/2018-2 - Universal Notice Data type The data that makes up this article is made up of several files used in its development, generated with the help of the software used in the development of the work, which made it possible to generate: Tables Images Graphics Maps Figures CSV SQL

Steps to reproduce

The development of the study, methods and protocols used were: Data Collection and Hardware and Software Development: - Initially, the study used research techniques and experiments to identify mechanisms that could offer solutions for the collection, storage and transmission of evapotranspiration data. - For this, factory calibrated digital sensors were used, adjustable through specific programming. DHT22 and DS18B20 sensors were chosen to measure temperature, humidity and transmit temperature data, respectively. Use of the ESP32 Microcontroller: - The ESP32 microcontroller was chosen due to its versatility and ability to integrate different technologies, such as Bluetooth, Wi-Fi and LoRa, making it suitable for various applications. Programming and Database: - The devices were programmed using the Arduino IDE, which offers a wide library of features and compatibility with various boards. - Furthermore, the Database used was based on SQL and configured with MySQL software, allowing flexibility of installation in different environments. Data Analysis and Comparison: - The collected data was processed and analyzed using software such as Power BI, RStudio and Matlab, and compared with records from a meteorological station. - The analyzes were carried out using regression and graphical visualizations, helping to understand the coefficients of determination and correlation of the data. This approach enabled the efficient collection, analysis and storage of data, contributing to the understanding and validation of the results obtained. Analysis applications: Willmott agreement index (d), determined by the equation: d=1-[(∑▒(Pi-Oi)^2 )/(∑▒(|Pi-O|+|Oi-O|)^2 )] “Where Pi corresponds to the values estimated by the Penman-Monteith methods, Oi to the estimated values, with O being the average of the estimates. Thus, “d” values can range from 0, for no agreement, to 1, for perfect agreement.” (Pereira et al., 2009, p. 2490). 〖ET〗_o=(0.408*∆*(R_N-G)+γ*900/(T+273)*U_2*(e_s+e_α ))/(∆+γ*(1+0.34*U_2 ) ) Where 〖ET〗_o the Reference Evapotranspiration in 〖day〗^(-1),R_N the Net Radiation in MJ*m^(-2) d^(-1) G the soil heat flux density in (taken equal to zero for the daily calculation step), ∆ the slope of the saturated vapor pressure curve in 〖k Pa〗^o C^(-1) , γ the psychometric coefficient in 〖k Pa〗^o C^(- 1) , u the wind speed measured at 2 m height in 〖m* sec〗^(-1) , e_s the saturated vapor pressure in kPa and e_α the real vapor pressure in kPa . Mean Bias Error (MBE) given by the formula: MBE=1/N ∑_(i=1)^N▒(x_(f,i)-x_(o,i) ) Root Mean Square Error systematic (〖RMSE〗_s) expressed in the formula: 〖RMSE〗_s=√(1/N ∑_(i=1)^N▒〖x_(f,i)-x_(o,i) ²〗) Root Mean Square Error unsystematic (〖RMSE〗_u) determined by: 〖RMSE〗_u=[(∑_(j=1)^N▒〖ω_j |P ̂_j-O_j |^2 〗)/(∑_(j=1)^N▒〖ω_j 〗)]^0.5 E Relative Root Mean Square Error (rRMSE) , determined by: rRMSE=[(∑_(j=1)^N▒〖ω_j |P ̂_j-P_j |^2 〗)/(∑_(j=1)^N▒〖ω_j 〗)]^0.5

Institutions

Categories

Funding

Licence