Prediction of reference evapotranspiration for irrigation scheduling using machine learning

Abstract The estimation of reference evapotranspiration (ET0) is important in hydrology research, irrigation scheduling design and water resources management. This study explored the capability of eight machine learning models, i.e., Artificial Neuron Network (ANN), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), Extreme Gradient Boosting (XGBoost), Multivariate Adaptive Regression Spline (MARS), Support Vector Machine (SVM), Extreme Learning Machine and a novel Kernel-based Nonlinear Extension of Arps Decline (KNEA) Model, for modeling monthly mean daily ET0 using only temperature data from local or cross stations. These machine learning models were also compared with the temperature-based Hargreaves–Samani equation. The results indicated that the estimation accuracy of these machine learning models differed in various scenarios. The tree-based models (RF, GBDT and XGBoost) exhibited higher estimation accuracy than the other models in the local application. When the station has only temperature data, the MARS and SVM models were slightly superior to the other models, while the ANN and HS models performed worse than the others. When there was no temperature data at the target station and the data from adjacent stations were used instead, MARS, SVM and KNEA were the suitable models. The results can provide a solution for ET0 estimation in the absence of complete meteorological data.

Download Full-text

Evapotranspiration data to determine agro-climatic zones in Egypt

Journal of Water and Land Development ◽

10.1515/jwld-2017-0009 ◽

2017 ◽

Vol 32 (1) ◽

pp. 79-86 ◽

Cited By ~ 20

Author(s):

Samiha A. H. Ouda ◽

Tahany A. Norledin

Keyword(s):

Nile Delta ◽

Reference Evapotranspiration ◽

Irrigation Scheduling ◽

Weather Data ◽

Climatic Zones ◽

Average Value ◽

Scheduling Model ◽

Significant Difference ◽

Difference Test ◽

Climatic Zoning

Abstract The objective of this paper was to compare between agro-climatic zones developed from 10-year interval of weather data from 2005-2014, 20-year interval of weather data from 1995-2014 and the zoning developed by [NORELDIN et al. 2016] using 30-year interval from 1985-2014 in the old cultivated land of Egypt in the Nile Delta and Valley. Monthly means of weather data were calculated for each year, and then monthly values for 10-year and 20-years were calculated for each governorate. Basic Irrigation scheduling model (BISm) was used to calculate reference evapotranspiration (ETo). Analysis of variance was used and the means was separated and ranked using least significant difference test (LSD0.05). Our results showed that agro-climatic zoning using 20-year values of ETo was similar to the zones developed with 30-year values of ETo, with different values of average ETo in each zone. Furthermore, using 10-year values of ETo resulted in higher values of ETo in each zone, compared to 20-year and 30-year ETo values. However, the average value of ETo over the three classifications was close to each other. Thus, depending on the availability of weather data, either zoning can be sufficient to develop agro-climatic zones.

Download Full-text

Investigation into the Effects of Climate Change on Reference Evapotranspiration Using the HadCM3 and LARS-WG

Water ◽

10.3390/w12030666 ◽

2020 ◽

Vol 12 (3) ◽

pp. 666 ◽

Cited By ~ 3

Author(s):

Maryam Bayatvarkeshi ◽

Binqiao Zhang ◽

Rojin Fasihi ◽

Rana Muhammad Adnan ◽

Ozgur Kisi ◽

...

Keyword(s):

Climate Change ◽

Reference Evapotranspiration ◽

Irrigation Scheduling ◽

Weather Data ◽

Base Period ◽

Resources Management ◽

Weather Information ◽

Daily Weather Data ◽

Weather Stations ◽

A1b Scenario

This study evaluates the effect of climate change on reference evapotranspiration (ET0), which is one of the most important variables in water resources management and irrigation scheduling. For this purpose, daily weather data of 30 Iranian weather stations from 1981 and 2010 were used. The HadCM3 statistical model was applied to report the output subscale of LARS-WG and to predict the weather information by A1B, A2, and B1 scenarios in three periods: 2011–2045, 2046–2079, and 2080–2113. The ET0 values were estimated by the Ref-ET software. The results indicated that the ET0 will rise from 2011 to 2113 approximately in all stations under three scenarios. The ET0 changes percentages in the A1B scenario during three periods from 2011 to 2113 were found to be 0.98%, 5.18%, and 12.17% compared to base period, respectively, while for the B1 scenario, they were calculated as 0.67%, 4.07%, and 6.61% and for the A2 scenario, they were observed as 0.59%, 5.35%, and 9.38%, respectively. Thus, the highest increase of the ET0 will happen from 2080 to 2113 under the A1B scenario; however, the lowest will occur between 2046 and 2079 under the B1 scenario. Furthermore, the assessment of uncertainty in the ET0 calculated by the different scenarios showed that the ET0 predicted under the A2 scenario was more reliable than the others. The spatial distribution of the ET0 showed that the highest ET0 amount in all scenarios belonged to the southeast and the west of the studied area. The most noticeable point of the results was that the ET0 differs from one scenario to another and from a period to another.

Download Full-text

Coupling a Bat Algorithm with XGBoost to Estimate Reference Evapotranspiration in the Arid and Semiarid Regions of China

Advances in Meteorology ◽

10.1155/2019/9575782 ◽

2019 ◽

Vol 2019 ◽

pp. 1-16 ◽

Cited By ~ 3

Author(s):

Yixiu Han ◽

Jianping Wu ◽

Bingnian Zhai ◽

Yanxin Pan ◽

Guomin Huang ◽

...

Keyword(s):

Air Temperature ◽

Reference Evapotranspiration ◽

Meteorological Data ◽

Bat Algorithm ◽

Irrigation Scheduling ◽

Multivariate Adaptive Regression Splines ◽

Accurate Estimation ◽

Agricultural Irrigation ◽

Semiarid Regions ◽

Arid And Semiarid

Accurate estimation of reference evapotranspiration (ETo) is key to agricultural irrigation scheduling and water resources management in arid and semiarid areas. This study evaluates the capability of coupling a Bat algorithm with the XGBoost method (i.e., the BAXGB model) for estimating monthly ETo in the arid and semiarid regions of China. Meteorological data from three stations (Datong, Yinchuan, and Taiyuan) during 1991–2015 were used to build the BAXGB model, the multivariate adaptive regression splines (MARS), and the gaussian process regression (GPR) model. Six input combinations with different sets of meteorological parameters were applied for model training and testing, which included mean air temperature (Tmean), maximum air temperature (Tmax), minimum air temperature (Tmin), wind speed (U), relative humidity (RH), and solar radiation (Rs) or extraterrestrial radiation (Ra, MJ m−2·d−1). The results indicated that BAXGB models (RMSE = 0.114–0.412 mm·d−1, MAE = 0.087–0.302 mm·d−1, and R2 = 0.937–0.996) were more accurate than either MARS (RMSE = 0.146–0.512 mm·d−1, MAE = 0.112–0.37 mm·d−1, and R2 = 0.935–0.994) or GPR (RMSE = 0.289–0.714 mm·d−1, MAE = 0.197–0.564 mm·d−1, and R2 = 0.817–0.980) model for estimating ETo. Findings of this study would be helpful for agricultural irrigation scheduling in the arid and semiarid regions and may be used as reference in other regions where accurate models for improving local water management are needed.

Download Full-text

Deep Reinforcement Learning-Based Irrigation Scheduling

Transactions of the ASABE ◽

10.13031/trans.13633 ◽

2020 ◽

Vol 63 (3) ◽

pp. 549-556

Author(s):

Yanxiang Yang ◽

Jiang Hu ◽

Dana Porter ◽

Thomas Marek ◽

Kevin Heflin ◽

...

Keyword(s):

Machine Learning ◽

Reinforcement Learning ◽

Real World ◽

Weather Conditions ◽

Irrigation Scheduling ◽

Water Application ◽

Time Step ◽

Net Return ◽

Crop Growth Stage ◽

Learning Machine

Highlights Deep reinforcement learning-based irrigation scheduling is proposed to determine the amount of irrigation required at each time step considering soil moisture level, evapotranspiration, forecast precipitation, and crop growth stage. The proposed methodology was compared with traditional irrigation scheduling approaches and some machine learning based scheduling approaches based on simulation. Abstract. Machine learning has been widely applied in many areas, with promising results and large potential. In this article, deep reinforcement learning-based irrigation scheduling is proposed. This approach can automate the irrigation process and can achieve highly precise water application that results in higher simulated net return. Using this approach, the irrigation controller can automatically determine the optimal or near-optimal water application amount. Traditional reinforcement learning can be superior to traditional periodic and threshold-based irrigation scheduling. However, traditional reinforcement learning fails to accurately represent a real-world irrigation environment due to its limited state space. Compared with traditional reinforcement learning, the deep reinforcement learning method can better model a real-world environment based on multi-dimensional observations. Simulations for various weather conditions and crop types show that the proposed deep reinforcement learning irrigation scheduling can increase net return. Keywords: Automated irrigation scheduling, Deep reinforcement learning, Machine learning.

Download Full-text

The GEFS-Based Daily Reference Evapotranspiration (ETo) Forecast and Its Implication for Water Management in the Southeastern United States

Journal of Hydrometeorology ◽

10.1175/jhm-d-13-0119.1 ◽

2014 ◽

Vol 15 (3) ◽

pp. 1152-1165 ◽

Cited By ~ 13

Author(s):

Di Tian ◽

Christopher J. Martinez

Keyword(s):

Water Management ◽

United States ◽

Reference Evapotranspiration ◽

Southeastern United States ◽

Demand Forecasting ◽

Skill Score ◽

Irrigation Scheduling ◽

Forecast Skill ◽

Ensemble Forecasts ◽

Relative Operating Characteristic

Abstract NOAA’s second-generation retrospective forecast (reforecast) dataset was created using the currently operational Global Ensemble Forecast System (GEFS). It has the potential to accurately forecast daily reference evapotranspiration ETo and can be useful for water management. This study was conducted to evaluate daily ETo forecasts using the GEFS reforecasts in the southeastern United States (SEUS) and to incorporate the ETo forecasts into irrigation scheduling to explore the usefulness of the forecasts for water management. ETo was estimated using the Penman–Monteith equation, and ensemble forecasts were downscaled and bias corrected using a forecast analog approach. The overall forecast skill was evaluated using the linear error in probability space skill score, and the forecast in five categories (terciles and 10th and 90th percentiles) was evaluated using the Brier skill score, relative operating characteristic, and reliability diagrams. Irrigation scheduling was evaluated by water deficit WD forecasts, which were determined based on the agricultural reference index for drought (ARID) model driven by the GEFS-based ETo forecasts. All forecast skill was generally positive up to lead day 7 throughout the year, with higher skill in cooler months compared to warmer months. The GEFS reforecast improved ETo forecast skill for all lead days over the SEUS compared to the first-generation reforecast. The WD forecasts driven by the ETo forecasts showed higher accuracy and less uncertainty than the forecasts driven by climatology, indicating their usefulness for irrigation scheduling, hydrological forecasting, and water demand forecasting in the SEUS.

Download Full-text

IRRIGATION SCHEDULING FOR PLASTIC GREENHOUSE MELON CROPS BASED ON HISTORICAL REFERENCE EVAPOTRANSPIRATION DATA

Acta Horticulturae ◽

10.17660/actahortic.2001.559.57 ◽

2001 ◽

pp. 383-389 ◽

Cited By ~ 1

Author(s):

M.D. Fernández ◽

S. Bonachela ◽

F. Orgaz ◽

E. Fereres

Keyword(s):

Reference Evapotranspiration ◽

Irrigation Scheduling

Download Full-text

Selecting models for the estimation of reference evapotranspiration for irrigation scheduling purposes

PLoS ONE ◽

10.1371/journal.pone.0245270 ◽

2021 ◽

Vol 16 (1) ◽

pp. e0245270

Author(s):

Lucas Borges Ferreira ◽

Fernando França da Cunha ◽

Sidney Sara Zanetti

Keyword(s):

Root Mean Square Error ◽

Reference Evapotranspiration ◽

Meteorological Data ◽

Irrigation Scheduling ◽

Soil Water Balance ◽

Suitable Model ◽

Mean Square ◽

Error Metrics ◽

Performance Gains ◽

Better Than

Alternative models for the estimation of reference evapotranspiration (ETo) are typically assessed using traditional error metrics, such as root mean square error (RMSE), which may not be sufficient to select the best model for irrigation scheduling purposes. Thus, this study analyzes the performance of the original and calibrated Hargreaves-Samani (HS), Romanenko (ROM) and Jensen-Haise (JH) equations, initially assessed using traditional error metrics, for use in irrigation scheduling, considering the simulation of different irrigation intervals/time scales. Irrigation scheduling was simulated using meteorological data collected in Viçosa-MG and Mocambinho-MG, Brazil. The Penman-Monteith FAO-56 equation was used as benchmark. In general, the original equations did not perform well to estimate ETo, except the ROM and HS equations used at Viçosa and Mocambinho, respectively. Calibration and the increase in the time scale provided performance gains. When applied in irrigation scheduling, the calibrated HS and JH equations showed the best performances. Even with greater errors in estimating ETo, the calibrated HS equation performed similarly or better than the calibrated JH equation, as it had errors with greater potential to be canceled during the soil water balance. Finally, in addition to using error metrics, the performance of the models throughout the year should be considered in their assessment. Furthermore, simulating the application of ETo models in irrigation scheduling can provide valuable information for choosing the most suitable model.

Download Full-text

Topological Sustainability of Crop Water Requirements and Irrigation Scheduling of Some Main Crops Based on the Penman-Monteith Method

Journal of Chemistry ◽

10.1155/2021/8552547 ◽

2021 ◽

Vol 2021 ◽

pp. 1-12

Author(s):

Waseem Asghar Khan ◽

Jamshaid Ul Rahman ◽

Mogtaba Mohammed ◽

Ziyad Ali AlHussain ◽

Murtada K. Elbashir

Keyword(s):

Reference Evapotranspiration ◽

Irrigation Scheduling ◽

Topological Indices ◽

Crop Water ◽

High Demand ◽

Crop Models ◽

Food And Agriculture ◽

Variable Topology ◽

Rain Fall ◽

Standard Software

The following method was used to apply the topology of the current study of evapotranspiration ETo, net irrigation demand, irrigation schedules, and total effective rain fall of different crop models: using the Food and Agriculture Organization's (FAO) CROPWAT 8.0 standard software and the CLIMWAT 2.0 tool and the FAO-56 Penman-Monteith approach to examine the variable topology of evapotranspiration ETo. Due to high temperatures in summer with an annual mean of 6.33 mm/day, the topological demonstration of reference evapotranspiration (ETo) increases from 2.84 mm/day in January to a maximum of 9.61 mm/day in July. Effective rainfall fluctuates from 0 mm to 53.4 mm. Total irrigation topological indices requirements were 308.3 mm/dec, 335.9 mm/dec, 343.6 mm/dec, 853 mm/dec, and 1479.6 mm/dec for barley, wheat, maize, rice, and citrus, respectively. The physical topological indices due to low demand in winter and high demand in summer, the total net irrigation, and gross irrigation for clay loamy soils for wheat (210.6 mm and 147.4 mm), barley (176.6 mm and 123.6 mm), citrus (204.5 mm and 143.2 mm), and maize (163.9 mm and 114.7 mm), but not for rice. This topology demonstrates that wheat has 4, barley has 4, citrus has 12, maize has 4, and rice crop has 12 irrigation schedules in a year.

Download Full-text