ArticlePublisher preview available

Daily Reference Evapotranspiration for Hyper-Arid to Moist Sub-Humid Climates in Inner Mongolia, China: I. Assessing Temperature Methods and Spatial Variability

Authors:
Xiaodong Ren at Inner Mongolia Agricultural University
Xiaodong Ren
Zhongyi Qu at Inner Mongolia Agricultural University
Zhongyi Qu
Diogo S Martins at University of Lisbon
Diogo S Martins
Paula Paredes at University of Lisbon
Paula Paredes
Show all 5 authorsHide
Read publisher preview
Download citation
To read the full-text of this research, you can request a copy directly from the authors.

Abstract and Figures

Abstract When weather data sets available for computing the reference evapotranspiration are incomplete or of questionable quality, there is the need to replace the FAO Penman- Monteith (PM-ETo) method by approaches requiring reduced sets only, particularly maximum and minimum temperature. The Hargreaves-Samani (HS) equation and the PM-ETo using only temperature data (PMT) are considered in this study and their results are compared with those of the PM-ETo using full datasets. Daily data sets refer to the period 1981–2012 and to a network of 50 meteorological stations covering the wide range of climates of Inner Mongolia. For both the PMT and HS methods, the solar radiation coefficients kRs were calibrated and have shown to be similar for both methods and to vary with climate aridity. For the PMT, the estimation of the dew point temperature (Tdew) was performed using the minimum temperature corrected for site aridity or, for humid climates, from a value near the average temperature. This improved estimation of Tdew was essential for a good performance of the PMT method in arid conditions and when temperatures are extremely low. RMSE <1 mm day−1 was obtained for both HS and PMT methods, and the modeling efficiency generally exceeded 0.85. The worse results correspond to windy and arid locations. The principal components analysis (PCA) in R-Mode have shown that the spatial variability of ETo computed with PM-ETo or with the HS and PMT methods were coherent. PCA supported the interpretation of ETo results. Overall, PMT performed better than HS for most locations. Keywords Grass reference evapotranspiration.Hargreaves-Samani(HS)eq. .PM-ETomethod. PMtemperaturemethod (PMT) . kRs radiationcoefficient . Principal component analysis .Aridity index
Figures - available from: Water Resources Management
This content is subject to copyright. Terms and conditions apply.
A preview of this full-text is provided by Springer Nature.
Learn more
Content available from Water Resources Management
This content is subject to copyright. Terms and conditions apply.
Xiaodong Ren
1
&Zhongyi Qu
1
&Diogo S. Martins
2,3
&
Paula Paredes
2
&Luis S. Pereira
2
Received: 12 November 2015 /Accepted: 31 May 2016 /
Published online: 18 June 2016
#Springer Science+Business Media Dordrecht 2016
Abstract When weather data sets available for computing the reference evapotranspiration
are incomplete or of questionable quality, there is the need to replace the FAO Penman-
Monteith (PM-ET
o
) method by approaches requiring reduced sets only, particularly maximum
and minimum temperature. The Hargreaves-Samani (HS) equation and the PM-ET
o
using only
temperature data (PMT) are considered in this study and their results are compared with those
of the PM-ET
o
using full datasets. Daily data sets refer to the period 1981–2012 and to a
network of 50 meteorological stations covering the wide range of climates of Inner Mongolia.
For both the PMT and HS methods, the solar radiation coefficients k
Rs
were calibrated and
have shown to be similar for both methods and to vary with climate aridity. For the PMT, the
estimation of the dew point temperature (T
dew
) was performed using the minimum temperature
corrected for site aridity or, for humid climates, from a value near the average temperature.
This improved estimation of T
dew
was essential for a good performance of the PMT method in
arid conditions and when temperatures are extremely low. RMSE <1 mm day
−1
was obtained
for both HS and PMT methods, and the modeling efficiency generally exceeded 0.85. The
worse results correspond to windy and arid locations. The principal components analysis
(PCA) in R-Mode have shown that the spatial variability of ET
o
computed with PM-ET
o
or
with the HS and PMT methods were coherent. PCA supported the interpretation of ET
o
results.
Overall, PMT performed better than HS for most locations.
Water Resour Manage (2016) 30:3769–3791
DOI 10.1007/s11269-016-1384-9
*Zhongyi Qu
quzhongyi@imau.edu.cn
*Luis S. Pereira
lspereira@isa.ulisboa.pt
1
Institute of Water Conservancy and Civil Engineering, Inner Mongolia Agricultural University,
Hohhot, Inner Mongolia 010018, China
2
LEAF, Instituto Superior de Agronomia, Universidade de Lisboa, Tapada da Ajuda, 1349-017 Lisboa,
Portugal
3
Instituto Dom Luiz, Faculdade de Ciências, Universidade de Lisboa, 1749-016 Lisboa, Portugal
Daily Reference Evapotranspiration for Hyper-Arid
to Moist Sub-Humid Climates in Inner Mongolia, China:
I. Assessing Temperature Methods and Spatial Variability
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
... However, Equation (2) was established for T dew = T min , and a correction for T dew is needed in practical applications [36]. Similarly, the adjustment coefficient k Rs in Equation (4) is an empirical value that differs between interior and coastal regions, and k Rs must also be corrected [11,[37][38][39][40]: ...
... Moreover, the solar radiation (R s ) can be calculated using Equations (9)- (11), and the actual saturated vapor pressure (e a ) can be estimated with Equation (2). However, the use of the PMF method requires correction for T dew [37][38][39][40]42]: (11) where N is the maximum possible duration of sunshine or daylight hours [hour]; n is the predicted sunshine duration [hour]; α is the sunshine duration coefficient; a s is the regression constant, reflecting the fraction of extraterrestrial radiation reaching the Earth on overcast days (n = 0); and a s + b s is the fraction of extraterrestrial radiation reaching the Earth on clear-sky days (n = N). Where no actual solar radiation data are available and no calibration is conducted to obtain improved a s and b s parameters, a s = 0.25 and b s = 0.50 are recommended [11,18,20,36]. ...
... The 1985 HS equation has been widely used in ET o estimation/prediction [9,11,22,23,[37][38][39][40][43][44][45][46][47][48][49]. The HS equation can be expressed as Equation (12), which has been applied in practice with corrections for parameters C and E [9,11,50,51]: ...
Evaluation of Empirical Equations and Machine Learning Models for Daily Reference Evapotranspiration Prediction Using Public Weather Forecasts
Article
Full-text available
  • Nov 2023
Although the studies on model prediction of daily ETo based on public weather forecasts have been widely used, these studies lack the comparative evaluation of different types of models and do not evaluate the seasonal variation in model prediction of daily ETo performance; this may result in the selected model not being the best model. In this study, to select the best daily ETo forecast model for the irrigation season at three stations (Yinchuan, Tongxin, and Guyuan) in different climatic regions in Ningxia, China, the daily ETos of the three sites calculated using FAO Penman–Monteith equations were used as the reference values. Three empirical equations (temperature Penman–Monteith (PMT) equation, Penman–Monteith forecast (PMF) equation, and Hargreaves–Samani (HS) equation) were calibrated and validated, and four machine learning models (multilayer perceptron (MLP), extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), and gradient boosting with categorical features support (CatBoost)) were trained and validated against daily observed meteorological data (1995–2015 and 2016–2019). Based on public weather forecasts and daily observed meteorological data (2020–2021), the three empirical equations (PMT, PMF, and HS) and four machine learning models (MLP, XGBoost, LightGBM, and CatBoost) were compared in terms of their daily ETo prediction performance. The results showed that the daily ETo performance of the seven models in the irrigation season with a lead time of 1–7 days predicted by the three research sites decreased in the order of spring, autumn, and summer. PMT was the best model for the irrigation seasons (spring, summer, and autumn) at station YC; PMT and CatBoost with C3 (Tmax, Tmin, and Wspd) as the inputs were the best models for the spring, autumn irrigation seasons, and summer irrigation seasons at station TX, respectively. PMF, CatBoost with C4 (Tmax, Tmin) as input, and PMT are the best models for the spring irrigation season, summer irrigation season, and autumn irrigation season at the GY station, respectively. In addition, wind speed (converted from the wind level of the public weather forecast) and sunshine hours (converted from the weather type of the public weather forecast) from the public weather forecast were the main sources of error in predicting the daily ETo by the models at stations YC and TX(GY), respectively. Empirical equations and machine learning models were used for the prediction of daily ETo in different climatic zones and evaluated according to the irrigation season to obtain the best ETo prediction model for the irrigation season at the study stations. This provides a new idea and theoretical basis for realizing water-saving irrigation during crop fertility in other arid and water-scarce climatic zones in China.
View
... Our study found that ET 0 in arid and semi-arid climate was more affected by air temperature, while solar radiation affected ET 0 more in humid and semi-humid climate (Figure 7). Several previous literature also found that the sensitivity of meteorological factors to ET 0 varied across climatic types [51][52][53][54]. When seasonal ET 0 was compared, the accuracy of ET 0 estimation decreased in autumn and winter seasons, especially for BP models in TC and PM zones. ...
... Our study found that ET0 in arid and semi-arid climate was more affected by air temperature, while solar radiation affected ET0 more in humid and semi-humid climate (Figure 7). Several previous literature also found that the sensitivity of meteorological factors to ET0 varied across climatic types [51][52][53][54]. When seasonal ET0 was compared, the accuracy of ET0 estimation decreased in autumn and winter seasons, especially for BP models in TC and PM zones. ...
Hybrid Genetic Algorithm−Based BP Neural Network Models Optimize Estimation Performance of Reference Crop Evapotranspiration in China
Article
Full-text available
  • Oct 2022
Precise estimation of reference evapotranspiration (ET0) is of significant importance in hydrologic processes. In this study, a genetic algorithm (GA) optimized back propagation (BP) neural network model was developed to estimate ET0 using different combinations of meteorological data across various climatic zones and seasons in China. Fourteen climatic locations were selected to represent five major climates. Meteorological datasets in 2018–2020, including maximum, minimum and mean air temperature (Tmax, Tmin, Tmean, °C) and diurnal temperature range (∆T, °C), solar radiation (Ra, MJ m−2 d−1), sunshine duration (S, h), relative humidity (RH, %) and wind speed (U2, m s−1), were first subjected to correlation analysis to determine which variables were suitable as input parameters. Datasets in 2018 and 2019 were utilized for training the models, while datasets in 2020 were for testing. Coefficients of determination (r2) of 0.50 and 0.70 were adopted as threshold values for selection of correlated variables to run the models. Results showed that U2 had the least r2 with ET0, followed by ∆T. Tmax had the greatest r2 with ET0, followed by Tmean, Ra and Tmin. GA significantly improved the performance of BP models across different climatic zones, with the accuracy of GABP models significantly higher than that of BP models. GABP0.5 model (input variables based on r2 > 0.50) had the best ET0 estimation performance for different seasons and significantly reduced estimation errors, especially for autumn and winter seasons whose errors were larger with other BP and GABP models. GABP0.5 model using radiation/temperature data is highly recommended as a promising tool for modelling and predicting ET0 in various climatic locations.
View
... For instance, the procedures by Hargreaves and Samani (1982), the equations proposed in FAO56 (Allen et al., 1998), the equations by Castellví et al. (1997), among others. However, these methods require local calibrations to obtain satisfactory performances as several studies have shown in various climate types ranging from hyper arid to humid (Karimi et al., 2020;Paredes et al., 2017;Raziei and Pereira, 2013;Ren et al., 2016;Todorovic et al., 2013). The approach of these studies was the calibration of the K RS coefficient of the Hargreaves and Samani (1982) method to estimate solar radiation when the variable was missing, and a correction in the minimum temperature to estimate actual vapor pressure in the absence of relative humidity data. ...
... However, Todorovic et al. (2013) found that for good ET o performance, the correction applied in the minimum temperature was only necessary for hyper-arid, arid, semi-arid, and dry subhumid climates and not for humid conditions. Moreover, Ren et al. (2016) found that the calibrated coefficient of the method to estimate solar radiation varied with climatic aridity. The errors of estimates were higher when the range of variation of ET o was higher, which occurred more often for hyper-arid and arid climates contrarily to sub-humid locations. ...
Improving reference evapotranspiration (ETo) calculation under limited data conditions in the high Tropical Andes
Article
  • Dec 2021
  • AGR WATER MANAGE
The computation of the reference crop evapotranspiration (ETo) using the FAO56 Penman-Monteith equation (PM-ETo) requires data on maximum and minimum air temperatures (Tmax, Tmin), relative humidity (RH), solar radiation (Rs), and wind speed (u2). However, the records of meteorological variables are often incomplete or of poor quality. Frequently, in the mountain areas such as those of the Andes, environmental sensors are subject to harsh conditions, due to the diurnal/nocturnal climatic variability causing challenging conditions for meteorological monitoring, which leads to data loss. For high-elevation landscapes like the Andes, the missing variables of vapor pressure deficit and solar radiation cause a high impact on PM-ETo calculation. To assess these limitations, several methods relying on maximum and minimum temperature to estimate the missing variables have been considered in the present investigation. Based on data from three automatic weather stations in the high Tropical Andes (humid páramo, 3298 – 3955 m a.s.l.), we found that the calibration and validation of methods were essential to estimate Rs. Using the (De Jong and Stewart, 1993) (Rs-DS) method we retrieved the highest performance, a RMSE between 2.89 and 3.81 MJ m⁻² day⁻¹. Moreover, In the absence of RH observations, replacing the dew point temperature (Tdew) by Tmin was a reliable alternative, when apply the method of (Allen et al., 1998) (VPD-FAO) which showed the highest performance with RMSE between 0.08 and 0.12 kPa. These results yielded highly accurate PM-ETo estimates, with RMSE between 0.29 and 0.34 mm day⁻¹ and RMSE between 0.12 and 0.18 mm day⁻¹, respectively. As expected, when both variables were missing, the ETo calculation increased its error, with an RMSE between 0.32 and 0.42 mm day⁻¹. A proper estimation of ETo in the Andean páramo contributes to improved water productivity for domestic and industrial uses, irrigated agriculture, and hydropower.
View
... The use of FAO56-PM ETo estimates as a standard for calibrating the Hargreaves equation has been previously utilized by other researchers from different regions (e.g., Hargreaves and Allen 2003;Hargreaves 1974;Gavin and Agnew 2003;Maeda et al. 2011;Cıtako et al. 2016). The Hargreaves equation is justified for the regions with sparse station networks, as it is costly to install lysimeters at all weather stations (Gavilan et al. 2006a;Gul et al. 2015;Ren et al. 2016;Djaman et al. 2018). In another study, data limitations in developing countries were tackled by utilizing five advanced machine learning algorithms, including additive regression variants, to predict monthly mean daily ETo in Pakistan's semi-arid region from 1987 to 2016 and for the different agro-climatic region of India (Elbeltagi et al. 2022a, b). ...
Short-Term Forecasting of Daily Reference Crop Evapotranspiration Based on Calibrated Hargreaves–Samani Equation at Regional Scale
Article
  • May 2024
This study focuses on enhancing real-time irrigation decisions and stream flow forecasts using short-term daily forecasts of reference evapotranspiration (ETo). While conventional approaches rely on historical observations for daily forecasting, developed countries have transitioned to issuing ETo forecasts derived from Numerical Weather Prediction or General Circulation Models (GCM) outputs. In this study, a similar approach was applied to predict short-term ETo forecasts for Pakistan using GCM output, combining data from the Pakistan Meteorological Department in-situ observation data and GCM forecast data from the Copernicus data source. ETo is calculated using the Hargreaves Samani (HS) equation, calibrated, and parameterized for the newly defined agro-climatic region by K-means clustering of ETo and soil moisture. Results indicate that the modified HS performs well in all climate regions except arid and humid regions, where errors are attributed to temperature forecast issues at high altitudes and the HS model neglecting wind speed and relative humidity effects. The integration of temperature data in the modified HS generates temperature-ETo correlation coefficients exceeding 0.92 in all agro-climatic regions. The method demonstrated accurate daily ETo forecasts for real-time irrigation predictions, particularly valuable in regions with sparse meteorological networks. The HS equation was calibrated for different agro-climatic regions using methods like fuzzy logic, pressure ratio, and curve fitting. The modified HS equation, integrated with numerical weather forecasts and a geographic information system, provides weekly forecasts and 10-day lead time ETo forecasts for district level and defined agro-climatic regions in Pakistan, with acceptable error ranges. Furthermore, the study determines a robust Pearson correlation (0.79) and a root-mean-square error of 0.54 with a 95% significant level, contributing significantly to predictive capabilities in the field of evapotranspiration forecasting.
View
... However, HS's equation overestimates (Mastrotheodoros et al., 2020) ET o at humid regions (Temesgen et al., 2005;Trajkovic, 2005) and underestimates ET o in very dry and windy regions (Fooladmand and Haghighat, 2007;Popova et al., 2006). Therefore, HS equation coefficient was locally calibrated and validated at 114 meteorological stations using linear regression between HS ET o and PM ET o and trial and error (TE) procedure (Raziei and Pereira, 2013;Ren et al., 2016). The corrected HS equation coefficient (C corr ) was used for ET o estimates with acceptable accuracy instead of the FAO-56 PM ET o equation. ...
Estimation of green and blue water evapotranspiration using machine learning algorithms with limited meteorological data: A case study in Amu Darya River Basin, Central Asia
Article
  • Dec 2022
Green-water evapotranspiration (GWET) and blue-water evapotranspiration (BWET) are much frequently discussed variables in the recent debates of water resources management and water productivity in water-scarce regions. But the deficiency of long-term, on-site records and limited observation stations is a critical challenge in determining the veracity of these variables. The GWET and BWET estimations rely considerably on extensive climate data, water fluxes data, soil parameters, crop distribution, and crop management data. However, obtaining accurate data by on-site observations or by remote sensing products is a difficult task in a data-scarce region and fewer variables are not sufficient to empirically estimate GWET and BWET. Machine learning (ML) is a modern artificial intelligence decision-making tool based on the analysis of fed-data and computer algorithms. This study reported the enormous potential of ML algorithms for estimating BWET and GWET using different sets of available climate variables. Wheat crop BWET and GWET were estimated at 114 meteorological stations in the Amu-Darya River Basin (ADRB) in Central Asia, using four most widely used ML algorithms: artificial neural network (ANN), supported vector machine (SVM), random forest (RF), and k-nearest neighbor (KNN). ML algorithms were trained with 75 % of the data, while tested and validated with 25 % of the data. A set of 24 models of different unique combinations of available variables were attempted to reasonably estimate GWET and BWET, and satisfying results were achieved. RF was found to be the most-promising ML algorithm to estimate BWET and GWET with limited available climate data. The estimated BWET and GWET can be considered in agriculture water resources policies to minimize further risks to the agroecosystem in ADRB.
View
... However, HS's equation overestimates (Mastrotheodoros et al., 2020) ET o at humid regions (Temesgen et al., 2005;Trajkovic, 2005) and underestimates ET o in very dry and windy regions (Fooladmand and Haghighat, 2007;Popova et al., 2006). Therefore, HS equation coefficient was locally calibrated and validated at 114 meteorological stations using linear regression between HS ET o and PM ET o and trial and error (TE) procedure (Raziei and Pereira, 2013;Ren et al., 2016). The corrected HS equation coefficient (C corr ) was used for ET o estimates with acceptable accuracy instead of the FAO-56 PM ET o equation. ...
Estimation of green and blue water evapotranspiration using machine learning algorithms with limited meteorological data: A case study in Amu Darya River Basin, Central Asia
Article
Full-text available
  • Jan 2022
Green-water evapotranspiration (GWET) and blue-water evapotranspiration (BWET) are much frequently discussed variables in the recent debates of water resources management and water productivity in water-scarce regions. But the deficiency of long-term, on-site records and limited observation stations is a critical challenge in determining the veracity of these variables. The GWET and BWET estimations rely considerably on extensive climate data, water fluxes data, soil parameters, crop distribution, and crop management data. However, obtaining accurate data by on-site observations or by remote sensing products is a difficult task in a data-scarce region and fewer variables are not sufficient to empirically estimate GWET and BWET. Machine learning (ML) is a modern artificial intelligence decision-making tool based on the analysis of fed-data and computer algorithms. This study reported the enormous potential of ML algorithms for estimating BWET and GWET using different sets of available climate variables. Wheat crop BWET and GWET were estimated at 114 meteorological stations in the Amu-Darya River Basin (ADRB) in Central Asia, using four most widely used ML algorithms: artificial neural network (ANN), supported vector machine (SVM), random forest (RF), and k-nearest neighbor (KNN). ML algorithms were trained with 75 % of the data, while tested and validated with 25 % of the data. A set of 24 models of different unique combinations of available variables were attempted to reasonably estimate GWET and BWET, and satisfying results were achieved. RF was found to be the most-promising ML algorithm to estimate BWET and GWET with limited available climate data. The estimated BWET and GWET can be considered in agriculture water resources policies to minimize further risks to the agroecosystem in ADRB.
View
... However, HS's equation overestimates (Mastrotheodoros et al., 2020) ET o at humid regions (Temesgen et al., 2005;Trajkovic, 2005) and underestimates ET o in very dry and windy regions (Fooladmand and Haghighat, 2007;Popova et al., 2006). Therefore, HS equation coefficient was locally calibrated and validated at 114 meteorological stations using linear regression between HS ET o and PM ET o and trial and error (TE) procedure (Raziei and Pereira, 2013;Ren et al., 2016). The corrected HS equation coefficient (C corr ) was used for ET o estimates with acceptable accuracy instead of the FAO-56 PM ET o equation. ...
Estimation of green and blue water evapotranspiration using machine learning algorithms with limited meteorological data: A case study in Amu Darya River Basin, Central Asia
Preprint
Full-text available
  • Oct 2022
  • COMPUT ELECTRON AGR
Green-water evapotranspiration (GWET) and blue-water evapotranspiration (BWET) are much frequently discussed variables in the recent debates of water resources management and water productivity in water-scarce regions. But the deficiency of long-term, on-site records and limited observation stations is a critical challenge in determining the veracity of these variables. The GWET and BWET estimations rely considerably on extensive climate data, water fluxes data, soil parameters, crop distribution, and crop management data. However, obtaining accurate data by on-site observations or by remote sensing products is a difficult task in a data-scarce region and fewer variables are not sufficient to empirically estimate GWET and BWET. Machine learning (ML) is a modern artificial intelligence decision-making tool based on the analysis of fed-data and computer algorithms. This study reported the enormous potential of ML algorithms for estimating BWET and GWET using different sets of available climate variables. Wheat crop BWET and GWET were estimated at 114 meteorological stations in the Amu-Darya River Basin (ADRB) in Central Asia, using four most widely used ML algorithms: artificial neural network (ANN), supported vector machine (SVM), random forest (RF), and k-nearest neighbor (KNN). ML algorithms were trained with 75 % of the data, while tested and validated with 25 % of the data. A set of 24 models of different unique combinations of available variables were attempted to reasonably estimate GWET and BWET, and satisfying results were achieved. RF was found to be the most-promising ML algorithm to estimate BWET and GWET with limited available climate data. The estimated BWET and GWET can be considered in agriculture water resources policies to minimize further risks to the agroecosystem in ADRB.
View
... The HS model has been widely used in various climates, and many studies have found that the values of the parameters C and E vary from region to region. To ensure model accuracy, the local calibration of these two parameters is necessary [32,34]. ...
Evaluation of Five Equations for Short-Term Reference Evapotranspiration Forecasting Using Public Temperature Forecasts for North China Plain
Article
Full-text available
  • Sep 2022
Accurate short-term forecasts of daily reference evapotranspiration (ET0) are essential for real-time irrigation scheduling. Many models rely on current and historical temperature data to estimate daily ET0. However, easily accessible temperature forecasts are relatively less reported in short-term ET0 forecasting. Furthermore, the accuracy of ET0 forecasting from different models varies locally and also across regions. We used five temperature-dependent models to forecast daily ET0 for a 7-day horizon in the North China Plain (NCP): the McCloud (MC), Hargreaves-Samani (HS), Blaney-Criddle (BC), Thornthwaite (TH), and reduced-set Penman–Monteith (RPM) models. Daily meteorological data collected between 1 January 2000 and 31 December 2014 at 17 weather stations in NCP to calibrate and validate the five ET0 models against the ASCE Penman–Monteith (ASCE-PM). Forecast temperatures for up to 7 d ahead for 1 January 2015–19 June 2021 were input to the five calibrated models to forecast ET0. The performance of the five models improved for forecasts at all stations after calibration. The calibrated RPM is the preferred choice for forecasting ET0 in NCP. In descending order of preference, the remaining models were ranked as HS, TH, BC, and MC. Sensitivity analysis showed that a change in maximum temperature influenced the accuracy of ET0 forecasting by the five models, especially RPM, HS, and TH, more than other variables. Meanwhile, the calibrated RPM and HS equations were better than the other models, and thus, these two equations were recommended for short-term ET0 forecasting in NCP.
View
... where AI is the site UNEP aridity index, and a T and a D are the correction factors, which vary and largely depend on local climate, as shown in Table 1 (Raziei and Pereira, 2013;Todorovic et al., 2013;Paredes and Pereira, 2019). This model has been proved appropriate to estimate ET 0 (e a data are needed) in various climate zones in many sites, such as Mediterranean countries (Todorovic et al., 2013), Inner Mongolia (Ren et al., 2016), Iran (Raziei and Pereira, 2013;Paredes et al., 2020), the Azores islands (Paredes et al., 2018), and continental Portugal (Paredes and Pereira, 2019). Equation (2a) with piecewise a T (depending on climate zones, Table 1), however, overestimated daily and monthly e a in hype-arid to semi-arid regions, although it greatly improved the accuracy compared to model without temperature correction. ...
Comparison of machine learning and dynamic models for predicting actual vapour pressure when psychrometric data are unavailable
Article
  • Jul 2022
  • J HYDROL
Information of actual vapour pressure (ea) is frequently required in many disciplines. However, psychrometric data required to calculate ea are often not readily available. Hence, it is of great importance to develop models to estimate ea when psychrometric data are unavailable. Here, five machine learning models were developed for estimating ea, viz. extreme gradient boosting (XGBoost), extreme learning machine (ELM), kernel-based nonlinear extension of Arps decline (KNEA), multiple adaptive regression splines (MARS), and support vector machine (SVM) models. Their performance was also compared to a dynamic model proposed recently, which estimates ea by adjusting dew point temperature from minimum temperature (Tmin) with dynamic correction factor. Three input combinations using only temperature data (i.e. Tmin and mean temperature (Tmean)) were considered in the machine learning models. The meteorological data collected from 1,188 stations across six climate zones were used to develop and assess the models. The overall results revealed that the dynamic and machine learning models offered satisfactory ea estimates spanning from hyper arid to humid climates. However, the accuracy of the dynamic model was lower than all machine learning algorithms using either only Tmin or combinations of Tmean and Tmin in all climate zones. The machine learning models using Tmean and Tmin were superior to those using only Tmean or Tmin. There were comparable performances among the ELM, KNEA, MARS, and SVM models with various input variables; however, the XGBoost model incorporating Tmean and Tmin produced the best accuracy. The computational demand was least for the ELM model, followed by the XGBoost model. Considering the accuracy and computational demand, the XGBoost model is recommended for predicting daily and monthly ea from hyper arid to humid climates when historical data are prior known. When there are no historical data, we recommend using the global XGBoost model incorporating Tmean, Tmin, and aridity index for estimating daily and monthly ea from arid to humid regions, and using the dynamic model in hyper-arid regions.
View
View
Daily Reference Evapotranspiration for Hyper-Arid to Moist Sub-Humid Climates in Inner Mongolia, China: II. Trends of ETo and Weather Variables and Related Spatial Patterns
Article
Full-text available
  • Sep 2016
  • WATER RESOUR MANAG
This study focuses on assessing trends of reference evapotranspiration (ETo) considering aridity. Weather data sets of 54–62 years of Inner Mongolia, a Chinese Province where climate varies from hyper-arid in the West to wet sub-humid in the East, were used. Trends were analyzed for ETo computed with the FAO Penman-Monteith method (PM-ETo) using full data sets of maximum and minimum temperature (Tmax and Tmin), sunshine duration (SD) used to compute net radiation, relative humidity (RH) and wind speed (WS). Trends were also assessed for ETo computed with the Hargreaves-Samani temperature eq. (ETo HS) and the Penman-Monteith equation with temperature estimates of solar radiation and actual vapour pressure (ETo PMT). In addition, trends relative to Tmax, Tmin, SD, RH and WS were assessed. Trends for PM-ETo show to vary with aridity, with decreasing trends in the areas marked by aridity in the West and increased trends in less arid and sub-humid areas in the East. The detected trends are well explained by the trends in weather variables which consist of large increasing trends of Tmax and Tmin and of decreasing trends for SD, RH and WS. Therefore, negative trends of ETo occur where impacts of increases in temperature and decreases in RH are smaller than impacts of declining SD and WS; otherwise, when warming influences are larger it results a positive trend for ETo. Trends were coherent when considering seasonality influences. Contrarily, results for the temperature methods, ETo PMT and ETo HS, always identified increased trends for ETo due to warming effects. These results show that it is inappropriate to assess ETo trends when using simplified temperature methods.
View
Symbolic Regression Equations for Calculating Daily Reference Evapotranspiration with the Same Input to Hargreaves-Samani in Arid China
Article
Full-text available
  • Apr 2016
  • WATER RESOUR MANAG
To present an alternative simple equation for reference evapotranspiration (ET o) estimation, the symbolic regression (SR) method was applied to establish equations with the same inputs to simple Hargreaves-Samani (HS) equation in arid China. For most of the equations derived by SR method for each station, their performance increased with an increase in the equation complex index (CI). The most precise equation performed well although it was always complex and greatly varied in form. On the other hand, the simplest one was uniform in equation structure and performed slightly better than the HS equation for all the five stations, and sometimes better than the local calibrated HS equation. A trade-off equation was selected with almost the same equation form for all the five stations and low CI index. The site-specific trade-off equation performed better than the simplest one and the locally calibrated HS equation. Then parameters in the trade-off equation were unified for all the five stations, it did not perform as good as the site-specific one, but performed better than the HS equation and unified local calibrated HS equation. Thus, the SR method is suitable to determine both the site-specific and the unified equation among stations for daily ET o calculation in arid regions.
View
Performance evaluation of reference evapotranspiration estimation methods, their interrelationships and recalibrated equations
Article
Full-text available
  • Jul 2015
Evapotranspiration is one of the major components of the hydrologic cycle and its accurate estimation is of paramount importance for hydrologic water balance studies, irrigation system design and management, crop yield simulation and water resources planning and management. A large number of indirect methods based on climatological data have been developed for the estimation of evapotranspiration. These methods vary from simple empirical relationships to complex methods. The choice of the methods, however, depends on its suitability in the region and available data. The present study aims to identify suitable empirical method for estimating evapotranspiration for the Tirupati and Nellore regions of Andhra Pradesh by comparing the measured evapotranspiration values with the values calculated by different prediction methods. The Penman-Monteith, corrected Penman and Pan Evaporation methods for groundnut crop, Penman-Monteith and Radiation methods for Red gram crop for Tirupati region, Hargreaves and Christiansen methods for Paddy for Nellore region were found suitable for estimating evapotranspiration. An attempt is also made to develop interrelationships among frequently used methods with FAO Penman-Monteith method so as to provide easy conversion facility among different methods. The influence of various climatic factors such as temperature, humidity, wind velocity, sunshine hours, vapour pressure and rain fall on daily crop evapotranspiration was studied using multiple correlation and regression analysis to develop relationships among them for different crops for these regions. The correlation analysis revealed that the influence of relative humidity, wind velocity, rainfall and vapour pressure during the monsoon season and temperature and sunshine during pre-monsoon is relatively more in the study regions on daily crop evapotranspiration. The relationships between evapotranspiration and influencing climatic parameters for groundnut, red gram and paddy crops for these regions were also developed.
View
View
Calculation methods for reference evapotranspiration with limited weather data (in Chinese)
Article
Full-text available
  • Mar 2001
Monthly data records of 40 Iranian stations distributed over the country, for the period 1971–2005, were utilized for estimation of reference evapotranspiration (ET o) using Penman–Monteith (PM-ET o), Hargreaves–Samani (HS) and FAO-PM temperature (PMT) methods. To estimate ET o with HS and PMT methods, appropriate k Rs , an empirical radiation adjustment coefficient, were considered for each station, whereas T min was adjusted for estimation of T dew and used only for PMT computation. It was found that the appropriate k Rs for both HS and PMT methods are identical for a given station and it is generally smaller in sub-humid and humid than in semi-arid to hyper-arid climates. The performance of the PMT was further improved in both arid and humid climates when T min was adjusted. The result suggested that the HS and PMT methods appropriately predict ET o for all climatic regions of Iran if the appropriate k Rs was utilized. However, the considered methods showed weak performances for some stations in arid and hyper-arid climates of eastern and southern Iran owing to the effect of extreme and variable wind speed inherent in the PM-ET o. Thus, the role played by wind speed in ET o estimation was examined; the result indicated that the existence of extreme winds, and also the time variability of wind speed, is responsible for the observed discrepancies between PMT and PM-ET o estimates. The spatial patterns of ET o computed with HS and PMT methods found to be identical and resemble to that of PM-ET o , all showing a gradual increasing from north to south, with the lowest ET o values observed over northern humid and sub-humid climates of Iran and larger ET o for arid and hyper-arid climates in the southern and eastern country. Results indicated that the HS and PMT methods are appropriate alternatives for estimation of ET o for all climatic regions of Iran.
View
Comparison of estimating ET0 with different methods in Hetao Irrigation District in Inner Mongolia
Article
  • Apr 2008
In order to obtain the proper calculation methods for reference crop evapotranspiration (ET0) in Hetao Irrigation District in Inner Mongolia and based on the micrometeorological data, four calculation methods (Priestly-Taylor. FAO Penman, Hargreaves-Samani, Irmark-Allen) for ET0 were compared with FA056 Penman-Monteith, respectively, and their applicabilities in Hetao Irrigation District were evaluated. The results show that calculation results of FAO Penman method is closest with FA056 Penman-Monteith, The average absolute error and average relative error are 0.43 mm/d and 12.50%, respectively. Other methods have different positive and negative deviations in different seasons. In the total calculation periods, deviations between the results by Irmark-Allen, Hargreaves-Samani and Penman-Monteith method are larger, so they can not be used in the district for its weather conditions, the results by FAO Penman method are similar with FA056 Penman-Monteith method and they can be applied to the most kinds of climate condition. Hargreaves-Samani method can be used in the district in which the difference in temperature is smaller. Irmark-Allen and Priestley-Taylor methods can be used in the area in which the relative humidity is larger.
View
View
View
View
Comparison between M5 Model Tree and Neural Networks for Estimating Reference Evapotranspiration in an Arid Environment
Article
  • Feb 2014
This paper describes a detailed evaluation of the performance and characteristic behaviour of feed-forward artificial neural network (ANN) and M5 model tree for estimating reference evapotranspiration (ET0) at four meteorological sites in an arid climate. The input variables for these models were the maximum and minimum air temperature, air humidity and extraterrestrial radiation. The FAO-56 Penman-Monteith model was used as a reference model for assessing the performance of the two approaches. The results of this study showed that the ANN estimated ET0 better than the M5 model tree but both models performed well for the study area and yielded results close to the FAO56-PM method. Root mean square error and R-2 for the comparison between reference and estimated ET0 for the tested data using the proposed ANN model are 5.6 % and 0.98, respectively. For the M5 model tree method these values are 8.9 % and 0.98, respectively. The overall results are of significant practical use because the temperature and Humidity-based model can be used when radiation and wind speed data are not available.
View