Chouaib El Hachimi

Timely and accurate estimations of reference evapotranspiration (ETo) are critical for effective water resource management and resilient agriculture planning in data-limited, climate-heterogeneous regions such as Africa. Here, we propose a physics-guided two-stage surrogate modeling framework to enable real-time estimation of daily ETo over Africa using satellite-derived surface solar radiation (Rs) and the TabPFN foundation model. First, a hybrid CNN-Windowed-Attention U-Net architecture was developed and trained using the spectral angle mapper loss to harmonize data from the MODIS Terra and Aqua sensors, thereby improving spatial coverage across Africa. Next, to capture radiative transfer processes, we build a surrogate model emulating MCD18A1.062 Rs from top-of-the-atmosphere reflectance and conditioning variables. Then, ETo was estimated in two ways: a physics-guided approach using surrogate Rs and a direct, end-to-end, data-driven approach. The framework was also evaluated against the widely adopted empirical methods, namely Priestley–Taylor, Makkink, Hargreaves–Samani, and Abtew, calibrated, in turn, against the FAO-56 Penman-Monteith as a reference. For Rs, results demonstrate that the surrogate model accurately reproduces Rs variability, with an RMSE of 2.03 MJ.m⁻².day⁻¹. For ETo, TabPFN improves from an RMSE of 0.93 mm.day⁻¹ in the end-to-end approach to 0.79 mm.day⁻¹ in the two-stage approach, outperforming both classical machine learning models and empirical formulations. Furthermore, continental and regional evaluations across nine IPCC climate regions show improved generalization of TabPFN across diverse hydroclimatic regimes, with RMSE values ranging from 0.26 to 0.91 mm.day⁻¹. In contrast, although they require difficult-to-obtain variables, empirical models exhibit systematic errors and limited adaptability to regional variability, with RMSE ranging from 0.49 to 2.94 mm.day⁻¹. With no reliance on costly ground measurements and being fully based on remotely sensed data, this framework can support timely estimation of irrigation water needs across Africa.

Smart management of weather data is an essential step toward implementing sustainability and precision in agriculture. It represents an important input for numerous tasks, such as crop growth, development, yield, and irrigation scheduling, to name a few. Advances in technology allow collecting this weather data from heterogeneous sources with high temporal resolution and at low cost. Generating and using these data in their raw form makes no sense, and therefore implementing adequate infrastructure and tools is necessary. For that purpose, this paper presents a smart weather data management system evaluated using data from a meteorological station installed in our study area covering the period from 2013 to 2020 at a half-hourly scale. The proposed system makes use of state-of-the-art statistical methods, machine learning, and deep learning models to derive actionable insights from these raw data. The general architecture is made up of four layers: data acquisition, data storage, data processing, and application layers. The data sources include real-time sensors, IoT devices, reanalysis data, and raw files. The data are then checked for errors and missing values using a proposed method based on ERA5-Land reanalysis data and deep learning. The resulting coefficient of determination (R²) and Root Mean Squared Error (RMSE) for this method were 0.96 and 0.04, respectively, for the scaled air temperature estimate. The MongoDB NoSQL database is used for storage thanks to its ability to deal with real-world big data. The system offers various services such as (i) weather time series forecasts, (ii) visualization and analysis of meteorological data, and (iii) the use of machine learning to estimate the reference evapotranspiration (ET₀) needed for efficient irrigation. To this, the platform uses the XGBoost model to achieve the precision of the Penman–Monteith method while using a limited number of meteorological variables (air temperature and global solar radiation). Results for this approach give R² = 0.97 and RMSE = 0.07. This system represents the first incremental step toward implementing smart and sustainable agriculture in Morocco.

Assessing the right amount of water needs for a specific crop is a key task for farmers and agronomists to achieve efficient and optimal irrigation scheduling, and then an optimal crop yield. To this end, the reference evapotranspiration (ET₀) was developed. It represents the atmospheric evaporation demand, and therefore an important variable for irrigation management. In this regard, several methods such as the FAO’s Penman-Monteith and Hargreaves have been used to model and estimate ET₀. These methods use climatic parameters data for calculation procedures such as solar net radiation (R_n), saturation vapour pressure(e_s), and min-max air temperatures or a combination of them. In this paper, we investigated two proposed data-driven methods to predict ET₀ values in a semi-arid region in Morocco. The first approach is based on forecasting techniques and the second one uses end-to-end modeling of ET₀ based on meteorological data and machine learning models. The feature selection and engineering results show that solar global radiation (R_g) and mean air temperature (T_a) have a significance of more than 87% as relevant predictors features for the ET₀. We then used them as input to machine learning regression models. Regression evaluation metrics showed that The XGboost regressor model performs well in both cross-validation with R²=0.93 in the first fold, and in hold-out validation with R²=0.92 and RMSE=0.55. As a final step, we compared the univariate time series forecasting of ET₀ using the Facebook Prophet model versus the machine learning modeling method that we proposed. As goodness-of-fit measures, forecasting using machine learning modeling of ET₀ showed better results in terms of both R² and RMSE.