Analysis and forecast of crop water demand in irrigation districts across the eastern part of the Ebro river basin (Catalonia, Spain): estimation of evapotranspiration through copernicus-based inputs

ORIGINAL PAPER Irrigation Science (2025) 43:637–654 https://doi.org/10.1007/s00271-024-00971-1 achieved by improving water management at farm and irrigation district (ID) levels (Playán & Mateos, 2005, Hsiao et al. 2007). In Spain, IDs are legal public corporations, attached to basin organizations, which are in charge of organizing the collective use of the public, surface and ground water that are common to them. Most irrigation water rights are collectively granted to all farmers operating within the same ID, organized into water user associations (GómezLimón et al., 2021). Water allocation varies between IDs, depending on the river basin, water storage reservoirs, level of modernization, and area to be irrigated. In addition, in periods of extreme drought, river basin managers consider water allocation and evaluate water use in irrigation areas (Rijsberman 2006). The traditional irrigation development paradigm is based on providing sufficient water to avoid water deficits at all times, so as to achieve maximum yields (Doorenbos and Pruitt 1992). Because of climate change and water supply constraints, this paradigm is changing (English et al. 2002; Bouman 2007) and quite often the allocation of irrigation Introduction Agriculture is the main user of freshwater, accounting for nearly 70% of total water consumption worldwide (Conforti 2011). Climate change, population growth and industrial development may decrease water availability exacerbating the challenge of meeting the demand for freshwater resources that already exist, especially in semi-arid Mediterranean regions (Bisselink et al. 2020). The sustainability of irrigated agriculture is strongly associated with the improvement of water use efficiency. This can be primarily Joaquim Bellvert [email protected] 1 Efficient Use of Water in Agriculture Program, Institute of AgriFood, Research and Technology (IRTA), Parc Agribiotech, Fruitcentre, Lleida 25003, Spain 2 Observatori de l’Ebre, Universitat Ramón Llull, CSIC, Roquetes 43520, Spain Abstract The agricultural sector is currently facing the uncertainty that accompanies climate change in terms of the availability of water resources, as well as the need to balance the water demand for agricultural irrigation with other uses in river basins. In Spain, irrigation districts (IDs) play a very important role in the management of water resources. The efficiency of ID water management involves finding an equilibrium between supply and demand. It is in relation to the latter where the uncertainty is greatest, because until now no tools have been available to characterize water demands with sufficient precision throughout irrigation campaigns. ID managers need precise information and the development of tools to support decision making in planning and water management. Therefore, this study aims to identify, compare and analyse the differences between the demands, allocations and consumptions of water for irrigation in different IDs of the eastern part of the Ebro basin during six consecutive growing seasons. In addition, projections of water demands up to 2100 are conducted using a dataset of six global climate models under different climate scenarios. Novel advances in remote sensing for evapotranspiration approaches using Copernicus-based inputs were used in this study. Large variabilities in water demands among IDs and in the adjustments between demands and allocations were observed, suggesting there is still much room for the improvement of water management. All climate projections have a very clear pattern indicating an upward trend in water demands until the end of the century. Received: 20 June 2024 / Accepted: 12 August 2024 / Published online: 27 August 2024 © The Author(s) 2024 Analysis and forecast of crop water demand in irrigation districts across the eastern part of the Ebro river basin (Catalonia, Spain): estimation of evapotranspiration through copernicus-based inputs JoaquimBellvert1· MagíPamies-Sans1· PereQuintana-Seguí2· JaumeCasadesús1 1 3 Irrigation Science (2025) 43:637–654 water is below the crop water requirements for maximum yield (Lorite et al. 2007). However, some IDs have water allocations much greater than their potential demands. This is usually associated with unmodernized IDs with very low water use efficiencies. This aspect is especially relevant and at the same time controversial, because in a context of water shortage, the excess water of one ID could be used by other IDs with lower water allocations but with more efficient irrigation systems. The choice of crop type to be planted depends to a certain extent on the annual water supplies, and in cases where there is a limited water allocation deficit irrigation strategies may be adopted. This is the case for certain IDs in the Ebro basin. Some of them, located in the southeastern part, such as Canals d’Urgell (CU) are still unmodernised and have an intermittent water supply for farmers, with an average turnaround of 15 days. Most of the farms within CU still use a flood irrigation system (Paolini et al. 2022). Others, such as Segarra-Garrigues (SG), have an on-demand irrigation scheme supplying water to the farm via pressurized pipe networks. Annual water allocation in the latter is fixed and varies from 1500 m3/ha to 6500 m3/ha, depending on the area. In contrast, the General Irrigators Community of the Canal of Aragon and Catalonia (CAYC) manages water for 105,000 ha, prorating it in accordance with supply and demand. In practice, this means that it allocates a maximum irrigation water amount every two weeks. Obviously, the definition of this apportionment is a critical factor since it should combine a priori antagonistic aspects. On the one hand, it must guarantee irrigation and enable agricultural production at all times, and on the other it must make it possible to ensure the availability of the water resource until the end of the irrigation campaign (Quintilla et al. 2014). Another very different case is the ID known as Garrigues Sud (GS), which only has a seasonal water allocation of 1500 m3/ha and whose farmers receive water only intermittently. In this scenario, the farms in this ID primarily grow olives and tree nut under sustained deficit irrigation. It is not enough to offer water to farmers as we need to improve the management of demand and promote efficiency in the use of this scarce resource. Knowledge of water availability is based on information about impounded volumes, current flows, historical records and data from snow reserve models. It is in relation to water demands where the uncertainty is greatest, because although there are historical records of the amount of water applied from previous years, there is no knowledge of the crops water needs in real time and their potential water consumption, making it impossible to characterize demands throughout the irrigation campaign. Therefore, ID managers require the development of tools to support decision making in planning and water management. Analysis of the temporal and spatial variability of water demands is necessary to encompass the diversity of irrigators and evaluate the operational and management decisions of each ID (van Opstal et al., 2022). In addition, the modelling of water demands in future climate change scenarios will be useful for ID managers to make decisions about changes in land use and the level of maximum water restrictions to adopt in conditions of water scarcity. Remote sensing technologies have the ability to estimate spatio-temporal water demands at ID scale and to analyse inter-annual variations. Crop evapotranspiration (ET) is a major component of the water balance and represents crop water requirements. Different studies have used the FAO-56 model to quantify a crop’s water consumption. This entails multiplying the reference ET of a well-irrigated grass with crop coefficients (Kc) (Allen et al. 1998), either estimated from linear relationships with vegetation indices (VIs) such as the normalized difference vegetation index (NDVI) or directly selected from the FAO-56 table (Casa et al. 2009; Er-Raki et al., 2007, Campos et al. 2010; Garrido-Rubio et al. 2020; Kharrou et al. 2021). Some IDs such as the CAYC have used this methodology to estimate crop water demands throughout the growing season (Casterad et al., 2015). It is important to note, however, that this approach depends on potential evapotranspiration (ETp) instead of actual evapotranspiration (ETa). The method is simple but, as crop coefficients only modulate the potential rate value yielded by a reference crop, this approach assumes that variations in ET between different crops are linear, which may not be the case for sparse woody crops as they differ widely from grass. A review of the advantages and disadvantages of VIbased crop coefficients was presented by Allen et al. (2011). For its part, the ETa can be derived from a soil-water-balance approach for which knowledge about the soil properties is needed to obtain the hydraulic properties through pedotransfer functions (PTFs) (Kharrou et al. 2021). However, one of the main limitations of this approach is the lack of detailed information regarding soil properties. Although some regions have very detailed maps of soil properties, a global high-resolution product is still lacking. Although the European Soil Database v2.0 is available at European level with a grid resolution of 1000 m (Hiederer 2013), when the aim is to estimate ETa through a soil-water-balance approach at field level and in drip-irrigated row crops, it could be argued that reliable soil water content data are often difficult to obtain because wet bulbs that develop below the emitters have a very heterogeneous pattern of soil moisture (Samadianfard et al. 2012). Recent advances in remote sensing for ET have been made through the use of surface energy balance (SEB) models (Bastiaanssen et al. 1998; Allen et al. 2007; Boulet et al. 2015; Norman et al. 1995). The great advantage of these models compared to optical methods such as crop coefficient 1 3 638 Irrigation Science (2025) 43:637–654 algorithms is that they can estimate ETa as opposed to ETp through the use of land-surface temperature (LST) measurements, thereby accounting for the influence of crop water stress (Senay et al. 2011; Knipper et al. 2019). Although extensive research has been carried out on ETa estimates, one of the major limitations of this approach continues to be the coarse spatial resolution of satellite thermal infrared images (Bellvert et al. 2020). While this is expected to improve in the near future with missions such as TRISHNA (Lagouarde et al. 2018) or the Sentinel high priority candidate LST mission (Koetz et al. 2018), several sharpening techniques have been developed to improve the spatio-temporal resolution of LST imagery (Gao et al., 2012, Semmens et al. 2016) to field scale. One of the applications that showcases this issue is the SEN-ET modelling framework (http://esa-sen4et.org) (Guzinski et al. 2020). This approach is based on the use of the two-source energy balance model (TSEB-PT) (Kustas et al. 2016) with Copernicus-based inputs. Validations in agricultural areas showed a RMSE of instantaneous latent heat flux of around 30%. Also, a recent study carried out in an almond orchard located in the Ebro basin (Spain) reported errors of 90 W/m2 and 87 W/m2 for the sensible (H) and latent heat flux (LE), respectively (Jofre-Čekalović et al. 2022). Research about remote sensing for ET has been mainly focused on improving methodologies for having a greater accuracy assessment. Although all modelling ET approaches still have room for improvement and many research efforts are going in that direction, the current maturity status allows it also to be applied as an operational system for monitoring and analysing water management of irrigation districts. This can be the case of the SEN-ET approach, which could operationally derive high-resolution daily ETa maps at 20 m resolution by sharpening thermal observations from Sentinel-3 satellites (1 km, daily) and optical observations from Sentinel-2 satellites (20 m, every 5 days). Therefore, the present study aims to take advantage of the SEN-ET modelling approach to quantify, compare, and analyse differences in crop water demands in several IDs located in the Ebro basin during six consecutive growing seasons (2017–2022). IDs are characterized by having different water allocations and regulations in water management, irrigation systems and crops. An in depth comparison analysis of crop water demand between years and IDs will give essential insights to water managers and can assist in decisions based on a better distribution of water allocations. In 2022, the whole of Europe suffered extreme drought conditions and, at least in Spain, some IDs were forced to reduce water allocations and apply restrictions. Therefore, particular emphasis is also made on studying how drought affected water demands over the course of 2022 compared to the previous years. Finally, projections of crop water demands in each ID are conducted in different climate scenarios (RCP4.5 and RCP8.5) up to 2100 using a dataset of six CMIP5 global climate models. This will contribute anticipate future trends in irrigation needs under climate change scenarios. Materials and methods Study area The study area is located in the irrigated area of Lleida (Catalonia, Spain), in the north-east of the Iberian Peninsula and inside the Ebro basin. Irrigation water management is organized by the following eight IDs: Canal d’Urgell (CU), Canal de Aragón y Cataluña (CAYC), Canal de Pinyana (CP), Segarra-Garrigues (SG), Algerri-Balaguer (AB), Segrià Sud (SS), Carrassumada (C) and Garrigues Sud (GS) (Fig. 1). In the case of GS, the area processed was 50% of the ID. With a total surface of ∼ 2000 km2, these IDs contain most of the Catalan irrigated surface area. This region includes major diversity among the IDs in terms of crops, irrigation systems, water allocations and agronomic practices. Table 1 summarizes some of the characteristics of each ID. Modernization has meant important changes in the irrigation system. Whereas in 1999 70% of the area was irrigated with surface irrigation, currently most of the irrigable area now has pressurized irrigation. The water comes from the Pyrenees and is stored in different reservoirs. The irrigation season starts when water is available for farmers, which usually occurs from mid-March, and continues through the end of October. Crop types vary among the IDs, with Fig. 2 showing the averaged percentages of crop distribution for the growing seasons 2017–2022. The modernization of some IDs, also with higher water allocations, has promoted double cropping. This consists of planting more than one crop per year, usually a combination of winter cereals followed by maize. The study area, which is located at altitudes ranging between 185 and 430 m a.s.l. has a typical semi-arid Mediterranean climate with mild, wet winters and very hot and dry summers. The average annual rainfall and reference evapotranspiration (ET0) are 300 mm and 1100 m, respectively. Remote sensing for ET In order to compute the crop water demand for each ID during the years 2017 to 2022, the two-source energy balance (TSEB) model was applied through Copernicus-based inputs. 1 3 639 Irrigation Science (2025) 43:637–654 soil, respectively. G is estimated as a fraction of the soil net radiation (Choudhury et al. 1987). G=cGRns (2) The fraction of net radiation which is stored in the soil, cG , is dependent on soil type, soil moisture and the time of the day, but typically is set as 0.35 for near-noon conditions (Norman et al. 1995). In the TSEB-PT model (Kustas and Norman 1999), an electrical circuit analogy is used in which H from soil and canopy are estimated based on three aerodynamic resistances to heat transport arranged in a series network. Since soil (Ts) and canopy temperature (Tc) cannot be directly retrieved from coarse resolution satellite-derived images, they are estimated in an iterative process in which The TSEB is a model formulated originally by Norman et al. (1995) and subsequently modified and improved by Kustas and Norman (1999). The TSEB models treat the land surface as two layers, soil and canopy, contributing to the energy and water fluxes (Eq. 1). Rn =Rnc+Rns (1a) Rnc=LEc+Hc (1b) Rns=LEs+Hs+G (1c) where Rn, H and LE respectively represents the partition of net radiation, the sensible and the latent heat of evaporation, with the subscripts c and s referring to canopy and Table 1 Summary of the total irrigated area, yearly water allocation and percentage of each irrigation system in each irrigation district (ID). Values corresponds to averaged numbers for the 2017–2021 period ID Total area (ha) Area of irrigated fields (ha) Water allocation (m3/ha) Irrigation system (%) Drip Sprinkler Flood AB 7881 6548 ∼ 6000 22.3 76.1 1.6 C 1351 1346 ∼ 7000 84.1 14.1 1.8 CAYC 29,955 29,635 ∼ 8000 33.0 65.1 1.9 CP 12,051 9223 ∼ 10,000 35.1 40.2 24.7 CU 64,088 64,333 ∼ 9000 9.9 5.4 84.7 GS 8748 7365 ∼ 1300 100 - - SG 68,328 25,383 1500–6500 82.5 15.4 2.1 SS 7687 6448 ∼ 2000 94.4 0.1 5.5 *Irrigation system percentages were obtained from Paolini et al. (2022), with the partial exception of CU which was modified according to the DUN-SIGPAC of year 2021 Fig. 1 Map of the study area in the northeast of Spain (Lleida, Catalonia) with the location of the different irrigation districts 1 3 640 Irrigation Science (2025) 43:637–654 The potential crop evapotranspiration (ETp) was also estimated using the Penman-Monteith equation (Allen et al. 2005). This method requires hourly meteorological data (solar irradiance, wind speed, vapour pressure deficit or relative humidity and air temperature), as well as certain information on crop conditions, such as albedo, LAI and/or stomatal resistance (Rc). In this study, crop parameters such as the LAI or albedo were directly obtained from Sentinel-2. The model assumes that whole canopy exchange can be adequately represented by simulating radiation capture and the partitioning of energy into H and LE as if it occurred at a single level, implying that there are no intermediate aerodynamic or convective resistances between vegetation and soil layers. This is known as the ‘big leaf’ approach. Copernicus input data sources The main satellite data inputs used in this study come from the Sentinel-2 (both A and B) and Sentinel-3 (both A and B). Sentinel-2 images at Level-2 A (L2A; bottom of atmosphere reflectance) were directly retrieved from CREODIAS (https://creodias.eu/), downloading four tiles for this study (T31[TBF, TCG, TBG, TCF]). In order to have a larger dataset of an agricultural area for the sharpening algorithm and to improve the accuracy of the machine learning model, the whole area of Lleida (north-east of Spain) (273940, 4573440, 359220, 4653320 m UTM 31 N) was it is first assumed that green canopy (expressed as the fraction of the leaf area index -LAIthat is green) transpires at a potential rate based on the Priestley-Taylor formulation (Priestley and Taylor 1972): LE c=αPTfg ∆ ∆+γ p Rnc (3) where ∆ is the slope of the saturated vapor pressure to the air temperature, γp is the psychometric constant, and fg is the fraction of the LAI that is green and therefore is transpiring. The αPT coefficient is set to 1.26. With this first estimate of LEc known, the canopy transpiration is then sequentially reduced (i.e., αPT <1.26) until realistic fluxes are obtained (LEc ≥ 0 and LEs ≥ 0). Further details on the TSEB-PT model scheme can be found at the source code (https://github.com/hectornieto/pyTSEB, DOI: https:// doi.org/10.5281/zenodo.594732, last accessed 20.08.2021) and in the original formulation of the model (Norman et al. 1995). The instantaneous latent heat flux (LE) (W.m-2) estimated from the TSEB-PT at the time of the thermal satellite overpass was upscaled to daily values of ETa (expressed in mm/day) by multiplying the instantaneous ratio of LE over solar irradiance by the average daily solar irradiance and assuming this ratio remains invariant during day-time hours (Cammalleri et al. 2014). Fig. 2 Maps of cumulative actual evapotranspiration (ETa) for the 2017–2022 growing seasons 1 3 641 Irrigation Science (2025) 43:637–654 only used at a time and place where thermal observations of the surface by the Sentinel-3 satellite were possible. In addition, solar irradiance was firstly estimated using AOT, TCWV and surface pressure and subsequently corrected by elevation, incidence angle, and terrain shading (Guzinski et al. 2021). Two ancillary data sources were used: Land cover maps from the Copernicus Climate Change Service (C3S) (https:// cds.climate.copernicus.eu/cdsapp#!/dataset/satellite-landcover?tab=overview, last accessed: 10.10.2022) and the Shuttle Radar Topography Mission (SRTM) DEM. In the study site, the CS3 landcover map, produced at 300 m resolution, did not vary from one year to another, and the one from 2017 was used for all years. A look-up table of different parameters associated to each crop class was generated following Guzinski et al. (2020). The data mining sharpening (DMS) approach (Gao et al., 2012) was used to sharpen the 1 km S3 LST imagery at 20 m resolution, as described in Guzinski et al. (2020). This methodology was previously used by Bellvert et al. (2020) and Jofre-Čekalović et al. (2022) in the same region. It relies on using temporal composites of S2 TOC reflectance at 20 m resolution centred on the date of S3 overpass. The pyDMS python module (https://github.com/radosuav/pyDMS last accessed 12.11.2021) was used to apply this algorithm. ET gap filling For dates with cloud coverage and therefore without a clear satellite image, a gap filling approach was applied. In this case, due to the S2 + S3 synergy methodology, it is possible to have occlusion either on S2 or S3 images. Therefore, in case one S2 data was not available, the previous and the following 5-day S2 images that were available were used. In case either the S3 image was cloud covered or S2 imagery was not available for more than 10 days, ETa for cloudy dates was retrieved from a crop stress coefficient (Kcs) obtained from adjacent dates and applied to the reference ET for cloudy dates (Jofre-Čekalović et al. 2022). Kcs was obtained on the assumption that the ratio of reference (FAO56) to actual ET remains steady over short periods. Crop classification analysis The DUN-SIGPAC database was used to classify crop types within each ID. DUN-SIGPAC contains information about crop type of parcels in Spain reported by farmers (as mandated by the European crop subsidy program) following the single agrarian declaration (Declaració Única Agrària, DUN) and provided annually by the Spanish Agricultural Land Geographic Information System (SIGPAC by its initials in Spanish) public administrative database. The database processed and mosaicked to retrieve the biophysical parameters required by TSEB-PT at 20 m spatial resolution. Since the 10 m resolution bands of S2 are not required in this study, band 8, which is a broader bandwidth version of band 8 A, was excluded. The S3 SLSTR sensor was used to obtain the thermal data. This includes 3 thermal infrared (TIR) channels with 1 km nominal spatial resolution and less than two days temporal resolution with one satellite at the equator and half a day when Sentinel 3 A and B are available (April 2018). The LST was obtained directly as an L2A product of S3 SLSTR at 1 km resolution. All the available images acquired in the 10 days ahead of a S2 acquisition in the first semester of 2017 and all SLSTR LST data with 5 days ahead from the second semester of 2017 to the end of 2022 were downloaded. A total of 35, 46, 45, 40, 35 and 44 S2 cloud-free images were downloaded for the period 15th March to 30th October respectively for 2017 to 2022. Likewise, 104, 101, 227, 228, 229 and 230 S3 cloud-free scenes were respectively fetched for the same years. From the 20-m top of canopy (TOC) S2 reflectance images, the biophysical parameters of the vegetation were computed through the Biophysical Processor (Weiss and Baret 2016) available in the SNAP software v.8.0 (https:// step.esa.int/main/download/snap-download/—last accessed 10.06.2021). The obtained biophysical parameters were the LAI, the fraction of vegetation cover (FVC), the fraction of absorbed photosynthetically active radiation (FAPAR), the canopy chlorophyll content (CCC) and the canopy water content (CWC). Python scripts were then used to estimate the fraction of vegetation that is green (fg – green LAI over total LAI), the vegetation gap fraction observed at the sensor viewing angle (fc (θ)), and leaf bi-hemispherical reflectance and transmittance, together with constant values for soil reflectance in the VIS-NIR to quantify the shortwave net radiation of the soil and canopy (Féret et al. 2017). Meteorological inputs were obtained from the European Centre of Medium Weather Forecast (ECMWF) ERA5 reanalysis dataset (Hersbach et al. 2020) and from the Copernicus Atmosphere Monitoring Service (CAMS) forecast dataset. The ERA5 data covers the whole Earth on a 30-km grid and with hourly temporal resolution. The data used from ERA5 consists of air temperature at 2 m, dew point temperature at 2 m, wind speed at 100 m, surface pressure, total column water vapour (TCWV) and surface geopotential. Aerosol optical thickness (AOT) at 550 nm was obtained from CAMS since it is not included in ERA5. A digital elevation model (DEM) was used to enhance the spatial resolution of air temperature, vapor pressure, and surface pressure and to correct variations due to changes in elevation at blending height (Guzinski et al. 2020). The final instantaneous parameter was surface solar irradiance. Clear sky conditions were assumed since this parameter is 1 3 642 Irrigation Science (2025) 43:637–654 Le Cointe 2022b), which provides daily gridded data of meteorological variables at 2.5 km of spatial resolution. An analogue resampling technique was used in order to reconstruct the new dataset of climate data on a daily basis for the period 2006–2100. This new dataset consists of a subset of six CMIP5 global climate models (GCMs) run for the Pyrenees region in Spain, based on CEDEX/MAPAMA (2017). This dataset was run in two contrasting scenarios: the RCP4.5 scenario, which represents the most optimistic scenario for CO2 emissions, and the RCP8.5, which represents the worst scenario in terms of CO2 emissions. The six CMIP5 global models were the following: MRI-CGCM3, MIROC-ESM, CNRM-CM5, MPI-ESM-MR, INMCM4 and BCC-CSM1-1. Then, the FAO-56 Penman-Monteith reference evapotranspiration (ET0) (Allen et al. 1998) was obtained for each model and the two contrasting scenarios. ET0 at 2.5 km were resampled at 20 m using the cubic spline method. Crop coefficients (Kc) maps were obtained for each date of the 2017–2021 growing period as the ratio of remotely sensed Penman-Monteith ETp and ET0. Daily averaged Kc for the five years were then used to obtain a new ETp for each date up to 2100 and for each of the assessed models and scenarios. Statistical analysis Differences between irrigation districts and crops were analyzed through a two-way analysis of variance (ANOVA) using the JMP Pro-Software (version 16.0, SAS Institute Inc., Cary, NC). Tukey’s HSD test was used to compare results when the source effect was significant, with p-values ≤ 0.05. Results and discussion Meteorological variables and statistical differences in ETa Meteorological variables such as reference ET (ET0) and rainfall showed a significant variability from year to year (Table 2). No statistical significant differences were observed in ET0 or rainfall between irrigation districts. ET0 was significantly higher in 2017, 2019 and 2022, averaging 1048 mm for the period studied. Years 2018 and 2022 respectively showed the highest and lowest rainfall values, with the latter 31% lower than the average of the previous five years. Cumulative ETa showed significant variations between irrigation districts and years. The interaction between ID and year was not significant. Cumulative ETa was significantly higher for 2017 and 2022, which on average accounted for contains several attributes for each parcel such as crop type, location and parcel identifier. Annual DUN-SIGPAC products were obtained from: http://agricultura.gencat.cat/ca/ serveis/cartografia-sig/aplicatius-tematics-geoinformacio/ sigpac/descarregues/. Also added recently was the irrigation system, but there are some discrepancies between SIGPAC inputs and reality (Paolini et al. 2022). In this study, eight main crop categories were selected for classification analysis: summer crops (mostly maize) [SC], double harvest crop (usually wheat/barley with maize or legumes with maize) [DC], fruit trees [FR], nut trees (almonds, walnuts and pistachios) [NT], olive trees [OLI], grapevines [GV], oilseed (i.e. sunflower or rapeseed) [OIL], and forage crops (mostly alfalfa) [FO]. In addition, horticultural crops [H] and others [OTH] were also added but not analysed. Accounting for irrigation district total water demands Total ID net (NWD) and gross water demand (GWD) were calculated (hm3) both daily and accumulated throughout each growing season, as: NWD = ETa – Precipitation (4). and. GWD = NWD/i.e. (5) where i.e. corresponds to irrigation system efficiency. Irrigation system maps developed by Paolini et al. (2022) in the area of study were used to calculate i.e. These maps were obtained by analysing temporal patterns of remotely sensed soil moisture and ETa using artificial intelligence algorithms and showed an accuracy of up to 90%. The coefficients used were 0.60, 0.75 and 0.90 for flood, sprinkler and drip irrigation, respectively (FAO 1989). Precipitation maps were also produced at 20 m resolution on a daily basis by interpolating data from all the 49 public agrometeorological stations available in the area (Dades meteorològiques de la XEMA | Dades obertes de Catalunya (transparenciacatalunya.cat). Firstly, hourly data was downloaded from all agrometeorological stations and transformed to daily precipitation. Then, the inverse distance weighting interpolation (IDW) approach was used to obtain daily precipitation maps at 20 m resolution, which were finally converted to annual precipitation maps. Modelling climate scenarios In order to generate projections of water demand in the study area, the PIRAGUA_atmos_climate dataset (https:// digital.csic.es/handle/10261/271116) was used (QuintanaSeguí and Le Cointe, 2022a). This dataset takes advantage of the already existing PIRAGUA-atmos_analysis (http:// digital.csic.es/handle/10261/271111) (Quintana-Seguí and 1 3 643 Irrigation Science (2025) 43:637–654 of these two IDs was the lowest. Finally, crops in SG were quite variable, since water allocation varies between zones. The main crops were double cropping (34%), followed by fruit, nut and olive trees (50%). Seasonal crop water consumption was analysed both together and for each ID separately (Table 3; Fig. 4). On average, forages, summer crops, double cropping and oilseeds had the significantly higher cumulative ETa, accounting for 891, 822, 849 and 834 mm/season, respectively. Grapevines, olive and nut trees had the lowest values, accounting for 682, 706, and 742 mm/season, respectively. The interaction between ID and crop type was not significant. In Fig. 4 the six-year averaged seasonal pattern of ETa for each crop can be observed by differentiating IDs. Summer crops followed a clear tend to increase throughout the growing season and reach maximum ETa values of 7 mm/ day around mid and late July. The double cropping (DC) pattern is also clear, where the two ETa peaks, one in spring and another in summer, can be observed. It can also be seen that ETa rates in fruit, nut and olive trees (FR, NT and OLI) and grapevines (GV) were lower in comparison to extensive crops (FO, OLE, OLE). Although there were no significant differences in the interaction between crop type and ID, it can be observed that ETa values of NT and OLI corresponding to the IDs of GS and SS were slightly lower than others, particularly during the summer period. This can be attributed to these two IDs having a lower water allocation than others, and hence the growers tending to adopt sustained deficit irrigation strategies. Similarly, FO and GV were also slightly lower in SG. We observed that values of ETa during winter and early spring, just before the start of vegetative growth, were higher than expected. This could be due to either the contribution of the herbaceous understory in non-tilled fields or cover crops in orchards/vineyards. Crop water demands at irrigation district level Figure 5 shows cumulative ETa, net water demand (NWD) and gross water demand (GWD) of each ID during the years 2017 to 2022. In total, averaged GWD for all Lleida’s irrigated area was 1036 ± 142 hm3. According to CHEBRO, the natural regime of the rivers Segre, Noguera Pallaresa and Ribagorçana, without the Cinca, is 3441 hm3 (CHEBRO, 2005). This means that the maximum amount of water needed through irrigation represents around 30% of the contributions in the natural regime. Cumulative ETa was always higher for CU due to its larger area, type of crops grown and irrigation systems (Fig. 5a). Total cumulative ETa in CU varied between 494 and 557 hm3, depending on the year. Table 1 indicates that the predominant irrigation system in CU is flood irrigation, 844 mm. On the other hand, 2019 and 2021 showed the lowest ETa values, with 729 and 750 mm, respectively. This means that except for 2019, ETa was linearly correlated with ET0. The shortages observed in 2019 may be due to the fact that it was a year with many heatwaves, especially during spring and early summer. This may have caused underirrigated fields to become water stressed and reduce their transpiration, therefore ETa. The irrigation districts with the highest ETa cumulative rates for irrigated fields were AB, CAYC and CU, followed by C and CP. The lowest ETa values were observed in GS and SS. Average accumulated ETa for the six studied years was 788 mm. Yearly differences of cumulative ETa between IDs can also be spatially observed in Fig. 2. Despite the significant differences in ETa between years, the spatial pattern remained constant throughout all years. Differences in water consumption between crop types Figure 3 shows the average percentages of each crop type in each irrigation district for the six evaluated years. Small interannual variation in the distribution of crop types with each ID were ignored. The predominant crops in AB, CU and CAYC were, in this order, double cropping, forage and fruit trees. These three together accounted for 70% of total crops. These three crops also represent 83% of CP, with fruit trees being the main one. C had 83% of fruit trees. Olive and nut trees were the main crops in GS and SS, with 97% and 78%, respectively. It should be noted that the irrigation allocation Table 2 Comparative analysis of reference evapotranspiration (ET0), rainfall and actual crop evapotranspiration (ETa) of irrigated parcels between irrigation districts (ID) for the period 15th March to 30th October. Different letters mean significant differences at p-value < 0.05 using Tukey’s honest significant difference test Source ET0 (mm) ETa (mm) Rainfall (mm) ID n.s < 0.0001* n.s Year 0.0039* < 0.0001* < 0.0001* ID x Year n.s n.s n.s Year ET0 (mm) ETa (mm) Rainfall (mm) ID ETa (mm) 2017 1050 a 854 a 263 c AB 929 a 2018 968 b 777 ab 450 a C 782 ab 2019 1042 a 729 b 281 c CAYC 889 a 2020 970 b 764 ab 377 b CP 823 ab 2021 974 b 750 b 250 c CU 859 a 2022 1051 a 833 a 223 c GS 607 c SG 734 b SS 682 c Mean 788 *Corresponds to significant differences at p ≤ 0.05; n.s, not significant 1 3 644 Irrigation Science (2025) 43:637–654 estimates of GWD are around 20% below the total water concession. In CU, double cropping, forage and summer crops respectively used 144 hm3, 143 hm3 and 116 hm3, followed by fruit trees which accounted for 90 hm3 (Fig. 6). CAYC was the second ID in extent and this was also reflected by its high crop water demands. Total cumulative ETa in CAYC varied between 273 and 245 hm3, depending on the year. In addition, NWD and GWD respectively ranged from 206 to 152 hm3 and from 264 to 194 hm3. Therefore, irrigation GWD of CAYC represents 23% of all Lleida’s irrigated area. Predominant water consumption in CAYC accounting for around 84% of fields. This means that most of the water applied is not used by crops but lost through evaporation or drainage. The effect of the low efficiency of this irrigation system is also reflected in the calculation of GWD, which clearly shows that CU accounts on average for all years for 520 ± 52 hm3 of water (Fig. 5c). This represents around 50% of the total GWD in Lleida’s irrigated area, which ranged between 849 hm3 (2018) and 1205 hm3 (2017). Since CU has an annual water concession of 630 hm3 distributed through two channels (the main one with 492 hm3 and the auxiliary with 138 hm3), it means that Fig. 3 Distribution of crop type within each irrigation district. Values are mean percentages of the six growing seasons. Irrigation districts correspond to Algerri-Balaguer (AB), Carrassumada (C), Canal Aragón y Cataluña (CAYC), Canal Pinyana (CP), Canals d’Urgell (CU), Garrigues Sud (GS), Segarra-Garrigues (SG), and Segrià-Sud (SS). Crop types correspond to summer crops (SC), double harvest crop (DC), fruit trees (FR), nut trees (NT), olive trees (OLI), grapevines (GV), oilseed (OIL), forage crops (FO), horticultural crops (H) and others (OTH) 1 3 645 Irrigation Science (2025) 43:637–654 (MICINN-AEI) from Spain. In addition, the work has indirectly received funding from the European Union’s Horizon Europe research and innovation programme under project ECO-READY (grant agreement No 101084201). Author contributions J.Bellvert wrote the main manuscript text. J.Bellvert and M.Pamies-Sans performed data processing and prepared figures. P. Quintana-Seguí provided data of climate projections. J.Casadesús contributed on the review of the manuscript and discussion. All authors reviewed the manuscript. Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. Data availability No datasets were generated or analysed during the current study. Declarations Competing interests The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not assessed differences in ETa and GWD in 2022 (year of high drought) in comparison to the period 2017–2021, and conducted projections of potential evapotranspiration (ETp) in the different IDs up to 2100 in two contrasting RCP scenarios. Results obtained showed that ETa in 2022 was 8% higher than the period 2017–2021 due to an increase in air temperature of 1.5ºC. This was also reflected in the GWD, which was 14% higher (568 m3/ha more). All climate projections have a very clear pattern indicating an upward trend until the end of the century. In conclusion, this study demonstrates the suitability of using the SEN-ET modelling approach to determine in near real-time the irrigation needs in large irrigation districts. The analysis conducted enriches the discussion on the impacts of climate change in the area. The results can also be used to guide the planning of targeted solutions and support the development of mitigation policies. In future projects, this methodology could be implemented as a decision-support tool in other irrigation districts. Acknowledgements This study was supported by the ALTOS project (PCI2019-103649) of the Partnership for Research and Innovation in the Mediterranean area (PRIMA) programme and by the ET4DROUGHT project (No. PID2021-127345OR-C31) of the Spanish Research Agency of the Ministry of Science and Innovation Fig. 10 Projections of Penman-Monteith potential crop evapotranspiration (ETp) calculated with six CMIP5 global climate models (GCM) under two different RCP scenarios (4.5 and 8.5) for each irrigation district. Bold lines correspond to the average for each RCP scenario 1 3 652 Irrigation Science (2025) 43:637–654 Choudhury BJ, Idso SB, Reginato RJ (1987) Analysis of an empirical model for soil heat flux under a growing wheat crop for estimating evaporation by an infrared-temperature based energy balance equation. Agric Meteorol 38(4):283–297 Confederación Hidrográfica del Ebro (2005) Implantación de la Directiva Marco. Caracterización de la demarcación y registro de zonas protegides [en linia]. http://www.chebro.es/contenido.visualizar. do?idContenido=4337&idMenu=3041. [Consulta: 26.05.2011] Conforti P (ed) (2011) Looking ahead in World Food and Agriculture: perspectives to 2050. Food and Agriculture Organization, Rome De la Fuente-Saiz D, Ortega-Farias S, Carrasco-Benavides M, OrtegaSalazar S, Tian F, Wang S, Liu Y (2024) Assessment of satellitebased water requirements for a drip-irrigated apple orchard in Mediterranean agroclimatic conditions. Heliyon 10(9):e29688 Doorenbos J, Pruitt WH (1992) Crop water requirements. FAO. Irrigation and drainage paper no. 24. Food and Agriculture. Organization of the United Nations, Rome English MJ, Solomon KH, Hoffman GJ (2002) A paradigma shift in irrigation management. J Irrig Drain Eng 128(5):267–277 Er-Raki S, Chehbouni A, Guemouria N, Duchemin B, Ezzahar J, Hadria R (2007) Combining FAO-56 model and ground-based remote sensing to estimate water consumptions of wheat crops in a semi-arid region. Agric Water Manage 87:41–54 FAO (1989) Food and Agriculture Organization of the United Nations. Irrigation Water management: Irrigation Scheduling. Training manual no. 4. https://www.fao.org/3/t7202e/t7202e00. htm#Contents Féret JB, Gitelson A, Noble S, Jacquemoud S (2017) PROSPECT-D: towards modeling leaf optical properties through a complete lifecycle. Remote Sens Environ 193:204–215 Gao F, Kustas WP, Anderson MC (2012) A Data Mining Approach for Sharpening Thermal Satellite Imagery over Land. Remote Sensing 4(11):3287–3319. Gómez-Limón JA, Gutiérrez-Martín C, Montilla-López NM (2021) Priority water rights. Are they useful for improving water-use efficiency at the irrigation district level?. Agricultural water management 257:107145 Garrido-Rubio J, Calera A, Arellano I, Belmonte M, Fraile L, Ortega T, Bravo R, González-Piqueras J (2020) Evaluation of remote sensing-based irrigation water accounting at river district management scale. Remote Sens 12:3187. https://doi.org/10.3390/ rs12193187 Guzinski R, Nieto H, Sandholt I et al (2020) Modelling High-Resolution actual evapotranspiration through Sentinel-2 and Sentinel-3 Data Fusion. Remote Sens 12(9):1433 Guzinski R, Nieto H, Sánchez JM et al (2021) Utility of Copernicusbased inputs for actual evapotranspiration modeling in support of Sustainable Water Use in Agriculture. IEEE J Sel Top Appl Earth Observations Remote Sens 14:11466–11484 Hersbach H et al (2020) The ERA5 global reanalysis, Quart J Roy Meteorological Soc 146(730):1999–2049. https://doi. org/10.1002/qj.3803 Hiederer R (2013) Mapping Soil properties for Europe - spatial representation of Soil Database attributes. EUR26082EN Scientific and Technical Research series, 47pp edn. Publications Office of the European Union, Luxembourg, pp 1831–9424. https://doi. org/10.2788/94128 Hsiao T, Steduto P, Fereres E (2007) A systematic and quantitative approach to improve water use efficiency in agriculture. Irrig Sci 25:209–231 Jofre-Čekalović C, Nieto H, Girona J, Pamies-Sans M, Bellvert J (2022) Accounting for Almond Crop Water Use under different irrigation regimes with a two-source Energy Balance Model and Copernicus-based inputs. Remote Sens 14:2106. https://doi. org/10.3390/rs14092106 included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/. References ACA (2023) https://aca.gencat.cat/ca/laigua/ estat-del-medi-hidric/recursos-disponibles/ estat-de-les-reserves-daigua-als-embassaments/ Allen RG (2005) Penman-Monteith equation. In: Hillel D (ed) Encyclopedia of soils in the environment. Elsevier, Oxford, pp 180–187 Allen RG, Pereira L, Raes D, Smith M (1998) Crop evapotranspiration: guidelines for Computing Crop Water requirements. FAO Irrigation and Drainage Paper 56. Food and Agriculture Organisation, Rome, Italy Allen RG, Tasumi M, Trezza R (2007) Satellite-based energy balance for mapping evapotranspiration with internalized calibration (METRIC)—Model. J Irrig Drain Eng Asce 133:380–394 Allen R, Pereira LS, Howell TA, Jensen ME (2011) Evapotranspiration information reporting: I. factors governing measurement accuracy. Agric Water Manage 98:899–920 Altés V, Bellvert J, Pascual M, Villar JM (2023) Understanding Drainage Dynamics and Irrigation Management in a Semi-arid Mediterranean Basin. Water 15:16. https://doi.org/10.3390/w15010016 Bastiaanssen WGM, Menenti M, Feddes RA, Holtslag AAM (1998) A remote sensing surface energy balance algorithm for land (SEBAL). 1. Formulation J Hydrol 212–213:198–212 Bellvert J, Jofre-Čekalović C, Pelechá A, Mata M, Nieto H (2020) Feasibility of using the two-source Energy Balance Model (TSEB) with Sentinel-2 and Sentinel-3 images to analyze the spatio-temporal variability of Vine Water Status in a Vineyard. Remote Sens 12:2299 Bisselink B, Bernhard J, Gelati E, Adamovic M, Guenther S, Mentaschi L et al (2020) Climate change and Europe’s water resources, EUR 29951 EN. Publications Office of the European Union, Luxembourg Boulet G, Mougenot B, Lhomme JP, Fanise P, Lili-Chabaane Z, Olioso A, Bahir M, Rivalland V, Jarlan L, Merlin O et al (2015) The SPARSE model for the prediction of water stress and evapotranspiration components from thermal infra-red data and its evaluation over irrigated and rainfed wheat. Hydrol Earth Syst Sci 19:4653–4672 Bouman BAM (2007) A conceptual framework for the improvement of crop water productivity at different spatial scales. Agric Syst 93:43–60 Cammalleri C, Anderson MC, Kustas WP (2014) Upscaling of evapotranspiration fluxes from instantaneous to daytime scales for thermal remote sensing applications. Hydrol Earth Syst Sci 18:1885–1894 Campos I, Neale CMU, Calera A, Balbontin C, González-Piqueras J (2010) Assessing satellite-based basal crop coefficients for irrigated grapes (Vitis vinifera L). Agric Water Manage 98:45–54 Casa R, Rossi M, Sappa G, Trotta A (2009) Assessing crop water demand by remote sensing and GIS for the Pontina Plain, Central Italy. Water Resour Manage 23:1685–1712 Casterad MA (2015) Aplicaciones basades en tecnologías de la información geogràfica para ayudar a gestionar el agua de riego en comunidades de regantes. de la Riva, J., Ibarra, P., Montorio, R., Rodrigues, M. (Eds.). Análisis espacial y representación geográfica: innovación y aplicación: 571–580. Universidad de Zaragoza-AGE. ISBN: 978-84-92522-95-8 1 3 653 Irrigation Science (2025) 43:637–654 Priestley CHB, Taylor RJ (1972) On the Assessment of Surface Heat Flux and Evaporation using large-scale parameters. Mon Weather Rev 100:81–92 Quintana-Seguí P, Le Cointe P (2022b) PIRAGUA_atmos_analysis [Dataset]; Observatori de l’Ebre (URL – CSIC); http:// digital.csic.es/handle/10261/271111; https://doi.org/10.20350/ digitalCSIC/14665 Quintana-Seguí P, Vidal J-P, Le Cointe P (2022a) PIRAGUA_atmos_ climate [Dataset]; Observatori de l’Ebre (URL – CSIC); http:// digital.csic.es/handle/10261/271116; https://doi.org/10.20350/ digitalCSIC/14666 Quintanilla-Albornoz M, Miarnau X, Pelechá A, casadesús J, GarcíaTejera O, Bellvert J (2023) Evaluation of transpiration in different almond production systems with two-source energy balance models from UAV thermal and multispectral imagery. Irrigation Science https://doi.org/10.1007/s00271-023-00888-1. Quintilla R, Callejero P, Casterad MA (2014) Apoyo a la gestión del agua en alta en la zona regable del Canal de Aragón y Cataluña con teledetección. XXXII Congreso Nacional de Riegos. Madrid, 11-12 June. http://hdl.handle.net/10261/98451 Van Opstal JD, Neale CMU, Hipps LE (2022) Evaluating the adaptability of an irrigation district to seasonal water availability using a decade of remotely sensed evapotranspiration estimates. Agricultural Water Management 26:107383. Rijsberman F (2006) Water scarcity: fact or fiction? Agric Water Manage 80(1–3):5–22 Samadianfard S, Sadraddini AA, Nazemi AH, Provenzano G, Kisi O (2012) Estimating soil wetting paterns for drip irrigation using genetic programming. Span J Agric Res 1155–1166 Semmens KA, Anderson MC, Kustas WP, Gao F, Alfieri JG, McKee L, Prueger JH, Hain CR, Cammalleri C, Yang Y, Xia T, Sanchez L, Alsina MM, Vélez M (2016) Monitoring daily evapotranspiration over two California vineyards using landsat 8 in a multi-sensor data fusion approach. Remote Sens Environ 185:155–170 Senay GB, Budde ME, Verdin JP (2011) Enhancing the simplified Surface Energy Balance (SSEB) approach for estimating landscape ET: validation with the METRIC model. Agric Water Manage 98:606–618 Stavros EN, Chrone J, Cawse-Nicholson K, Freeman A, Glenn NF, Guild L, Kokaly R, Lee C, Luvall J, Pavlick R, Poulter B, Schollaert Uz S, Serbin S, Thompson DR, Townsend PA, Turpie K, Yuen K, Thome K, Wang W, Zareh S-K, Nastal J, Bearden D, Miller CE, Schimel D (2023) Designing an observing system to study the surface biology and geology (SBG) of the earth in the 2020s. J Geophys Res: Biogeosci 128 (1), https://doi. org/10.1029/2021JG006471, URL https://onlinelibrary.wiley. com/doi/10.1029/2021JG006471 Weiss M, Baret F (2016) S2ToolBox Level 2 products: LAI, FAPAR, FCOVER—Version 1.1. Sentin. ToolBox Level2 Prod. 53. Available online: https://step.esa.int/docs/extra/ATBD_S2ToolBox_ L2B_V1.1.pdf (accessed on 19 May 2021) Zheng Y, Hao L, Shedayi AA, Guo L, Ma C, Huang B, Xu M (2018) The optimal CO2 concentrations for the growth of three perennial grass species. BMC Plant Biol 18:27. https://doi.org/10.1186/ s12870-018-1243-3 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Kharrou MH, Simonneaux V, Er-Raki S, Le Page M, Khabba S, Chehbouni A (2021) Assessing Irrigation Water Use with Remote sensing-based Soil Water Balance at an Irrigation Scheme Level in a Semi-arid Region of Morocco. Remote Sens 13:1133. https:// doi.org/10.3390/rs13061133 Kimball BA, Manuney JR, Nakayama FS, Idso SB (1993) Effects of increasing atmospheric CO2 on vegetation. Vegetatio 104:65–75. https://doi.org/10.1007/BF00048145 Knipper KR, Kustas WP, Anderson MC, Alfieri JG, Prueger JH, Hain CR, Gao F, Yang Y, McKee LG, Nieto H et al (2019) Evapotranspiration estimates derived using thermal-based satellite remote sensing and data fusion for irrigation management in California vineyards. Irrig Sci 37:431–449 Koetz B, Bastiaanssen W, Berger M, Defourney P, Bello UD, Drusch M, Drinkwater M, Duca R, Fernandez V, Ghent D, Guzinski R, Hoogeveen J, Hook S, Lagouarde J, Lemoine G, Manolis I, Martimort P, Masek J, Massart M, Notarnicola C, Sobrino J, Udelhoven T (2018) High spatiotemporal resolution land surface temperature mission - A copernicus candidate mission in support of agricultural monitoring. In: IGARSS 2018–2018 IEEE International Geoscience and Remote Sensing Symposium. 8160–8162. https://doi.org/10.1109/IGARSS.2018.8517433 Kustas WP, Norman JM (1999) Evaluation of soil and vegetation heat flux predictions using a simple two-source model with radiometric temperatures for partial canopy cover. Agric Meteorol 94:13–29 Kustas WP, Nieto H, Morillas L, Anderson MC, Alfieri JG, Hipps LE, Villagarcía L, Domingo F, García M (2016) Revisiting the paper “using radiometric surface temperature for surface energy flux estimation in mediterranean drylands from a two-source perspective”. Remote Sens Environ 184:645–653 Lagouarde J-P, Bhattacharya B, Crebassol P, Gamet P, Babu SS, Boulet G, Briottet X, Buddhiraju K, Cherchali S, Dadou I, Dedieu G, Gouhier M, Hagolle O, Irvine M, Jacob F, Kumar A, Kumar KK, Laignel B, Mallick K, Murthy C, Olioso A, Ottle C, Pandya MR, Raju PV, Roujean J-L, Sekhar M, Shukla MV, Singh SK, Sobrino J, Ramakrishnan R The Indian-French Trishna mission: Earth observation in the thermal infrared with high spatio-temporal resolution. In: IGARSS 2018–2018 IEEE International Geoscience and Remote Sensing Symposium. IEEE, Valencia (2018) 4078–4081. https://doi.org/10.1109/IGARSS.2018.8518720, URL https://ieeexplore.ieee.org/document/8518720/ Le Page M, Fakir Y, Jarlan L, Boone A, Berjamy B, Khabba S, Zribi M (2021) Projection of irrigation water demand based on the simulation of synthetic crop coefficients and climate change. Hydrology Earth Syst Sci 25:637–651. https://doi.org/10.5194/ hess-25-637-2021 Lorite IJ, Mateos L, Orgaz F, Fereres E (2007) Assessing deficit irrigation strategies at the level of an irrigation district. Agric Water Manage 91:51–60 Norman JM, Kustas W, Humes K (1995) A two-source approach for estimating soil and vegetation energy fluxes from observations of directional radiometric surface temperature. Agric Meteorol 77:263–293 Paolini G, Escorihuela MJ, Merlin O, Pamies-Sans M, Bellvert J (2022) Classification of different Irrigation systems at Field Scale using Time-Series of Remote Sensing Data. IEEE J Sel Top Appl Earth Observations Remote Sens 15:10055–10072. https://doi. org/10.1109/JSTARS.2022.3222884 Playán E, Mateos L (2005) Modernization and optimization of irrigation systems to increase water productivity. Agric Water Manage 80:100–116 1 3 654