Digital twin (DT) technology is attracting increasing interest as a potentially valuable tool for the future of agriculture. By offering a dynamic virtual representation of real agricultural systems, it opens up new possibilities for real-time monitoring, simulation, and decision support. In principle, such approaches could improve predictive capacity, optimize resource use, and support more responsive management strategies. However, agriculture cannot be treated as an engineered system, and this is where important challenges emerge. Agroecosystems are living, context-dependent, and inherently variable, shaped by diverse processes that remain only partly observable and often difficult to model. This makes their representation and prediction considerably more complex than in many industrial applications. In this review, we critically examine the conceptual foundations, architectural frameworks, and current applications of agricultural digital twins (ADTs), while also identifying key scientific and practical constraints that continue to limit their development. Particular attention is given to two recurring issues: the assumption that increasing data availability necessarily improves prediction, and the persistent gap between observable variables and the underlying biological and ecological processes that govern system behaviour. Drawing on conceptual figures and comparative analyses, we highlight important research gaps and argue for a shift in perspective. Rather than pursuing increasingly precise predictions, there is a need to develop digital twins that explicitly account for uncertainty and support more resilient forms of decision-making. In this context, the value of ADTs may lie less in predictive accuracy alone, and more in their ability to help decision-makers navigate complexity, variability, and change.

Accurate rainfall-streamflow observations are essential for understanding runoff generation and flood hazards in mountainous regions. Yet such information remains scarce in many areas, especially for small and medium-sized catchments in tropical environments of developing countries. This study develops high-resolution rainfall-streamflow datasets and evaluates the spatiotemporal variability of rainfall-runoff dynamics to assess the applicability of the rational method, from its empirical formulation to its physical interpretation, for flood-hazard management in the Nyamutera (44 km2) and Gaseke (109 km2) catchments of the mountainous Mukungwa watershed in northwestern Rwanda. To address data scarcity, cost-effective stream monitoring stations were installed to record flow depths, complemented by three automatic rain gauges and two weather stations that record rainfall and other weather parameters at 15-min intervals. Periodic discharge measurements were conducted to establish stage-discharge rating curves, which were used to derive continuous streamflow records from April 2022 to May 2023. Analysis of 40 storm-event responses revealed marked contrasts between the catchments: Nyamutera produced higher runoff coefficients (0.05-0.40; mean = 0.20; annual mean = 0.18) than Gaseke (0.02-0.35; mean = 0.10; annual mean = 0.11), reflecting its steeper slopes and lower storage capacity. Design runoff coefficients for the 100-year event were 0.55 and 0.51, and recorded peak discharges reached 131 and 122 m3/s, respectively. These values reflect Nyamutera's faster, more concentrated flow pathways, in contrast to the more attenuated response observed in Gaseke. The results highlight differences in runoff behaviour between the two catchments, likely reflecting their contrasting topography, soil properties and storage capacity. Although the monitoring network captured most dynamics, uncertainties remain in monitoring very low flows, extreme peaks and high rainfall variability. This study suggests additional monitoring techniques that could help capture these conditions. The developed approach provides a practical, transferable framework for rainfall-streamflow characterisation in similarly data-limited mountainous environments.

Lake Guidimouni, a designated Ramsar wetland in Niger’s Zinder Region, represents a keystone socio-ecological system whose ecological functions, cultural heritage, and local livelihoods are tightly interwoven. Yet, this emblematic dryland lake is increasingly destabilised by intensifying climatic extremes—prolonged droughts and recurrent floods—and by accelerating anthropogenic pressures. To provide a more rigorous and climate-centred diagnosis of these transformations, we developed and operationalized a spatially explicit, participatory and statistically validated DPSIR–GIS framework integrating object-based image analysis, spatial autocorrelation modelling and NDVI-based carbon proxies. This combined approach enabled the development of a fine-scale vulnerability map of the lake and its periphery, capturing both biophysical and socio-economic dimensions of change. Our findings reveal two converging threats jeopardising ecosystem sustainability: (i) extensive inundation of agricultural lands and (ii) the massive proliferation of the invasive macrophyte Typha domingensis. Land-use analysis demonstrates that approximately 77.2 ha of irrigated croplands have transitioned into aquatic grasslands, signalling a structural ecological shift tightly linked to hydrological stress and climate variability. Moreover, 42.5% of irrigated agriculture is now exposed to flooding and exhibits high susceptibility to vegetative invasion. In response to these pressures, we identify targeted, science-based strategies—such as selective mechanical harvesting, hydrological regulation, and adaptive local governance—as priority resilience levers capable of restoring ecological balance and safeguarding ecosystem services. These findings underline the urgent need for integrated natural-resource management in Sahelian drylands. The analysis further integrates convergent stakeholder perceptions collected through participatory co-production, which were systematically embedded within the DPSIR framework to triangulate and validate spatial and remote-sensing evidence.

The world faces major environmental challenges, including climate change, soil degradation, water scarcity, and pollution from unmanaged organic waste. Sustainable and scientifically validated strategies are therefore essential. This study evaluated the effect of a compost-vermicompost mixture (70:30, w/w), derived from food waste produced by the Vert d'Iris Cooperative (Belgium), on soil field capacity and humic acid quality. The amendment was produced and characterized according to international standards before application to the Betteraves Enz garden (Anderlecht). Soil samples collected at 0-30 cm depth from amended plots (n = 15, n = number of samples) and control plots (n = 15) were analyzed for physicochemical properties and water retention. Humic acids were extracted following the International Humic Substances Society protocol and characterized by UV-visible and FTIR spectroscopy. Amended soils showed a significantly higher field capacity than controls (38.4% vs. 27.3%), representing an increase of approximately 40% (p < 0.001, i.e., probability of error below 0.1%). Spectroscopic analyses indicated enhanced humification, reflected by significant variations (p < 0.05) in the absorbance ratios A2/4 (E2/E4, 280/465 nm) and A4/6 (E4/E6, 465/665 nm), as well as in the Delta logK index (logA400 - logA600), which are indicators of aromatic condensation, molecular size, and structural maturity of humic substances. FTIR spectra confirmed enrichment in aromatic and oxygenated functional groups. These findings demonstrate that combining physicochemical and spectroscopic indicators provides robust evidence of improved soil water retention and long-term organic matter stabilization, supporting sustainable soil management strategies.

Fertiliser microdosing (FM) is widely promoted in sub-Saharan Africa as a means to improve fertiliser use efficiency in cereal crops. However, combining organic amendments with FM could further increase the agronomic efficiency of the latter because of possible synergies between both practices. This study therefore evaluated whether combining FM with organic amendments enhanced fertiliser use efficiency in sorghum crops and how local biophysical or management factors mediate this interaction. Over 2 years, 225 on-farm trials were conducted across three provinces of Burkina Faso along a north-south climatic gradient. In the northern (Zandoma) and central (Oubritenga) provinces, eight treatments were tested at each site: FM (N-P-K 14-23-14), hill-placed compost, hill-placed compost + FM, hill-placed compost with rock phosphate (compost-BP), hill-placed compost-BP + FM, hill-placed biochar, hill-placed biochar macerated with fertiliser, and control plots. In the Sissili province (south), compost-BP was replaced by N-P-K 23-10-05. The highest sorghum yields were observed when FM was combined with compost or compost-BP. However, yield gains attributable to FM (N-P-K 14-23-14) were not improved by adding compost, compost-BP, or biochar compared to FM alone. In Sissili, FM with N-P-K 23-10-05 outperformed the conventional 14-23-14 formulation, highlighting the need to adapt fertiliser blends to agro-climatic zones. Across all treatments, crop response to FM was strongly soil type-dependent. In Zandoma, the driest province, it was observed that the residual effects of previous fertiliser applications affected yield response to FM. The results suggest that hill-placement of macerated biochar or the combination of FM with compost-BP are not suitable for optimising smallholder sorghum production. Combining FM with compost is recommended where sufficient compost is available. Otherwise, applying FM and compost in different fields or parts of fields is to be recommended. Further research is needed to explain the poor response to macerated biochar in Oubritenga and Zandoma.

Crop simulation models are essential decision support tools, particularly in regions with highly variable climates, as they help farmers optimize crop management practices. In Moroccan rainfed areas, nitrogen (N) and phosphorus (P) fertilization strategies are critical for making informed decisions about the timing and quantity of application to maximize income. This research presents an innovative rule-based approach that integrates the APSIM crop simulation model to enhance nitrogen (N) and phosphorus (P) fertilization recommendations, tailored specifically for implementation at farmers’ field levels. This study compares four N- and P-fertilization strategies: two rule-based strategies derived from the APSIM model using historical climate data and current-year rainfall, and two recommendations from empirical models commonly used by Moroccan farmers and advisers (Fertimap and NPK-engine). The study tested the impact of these four fertilization schemes on wheat yield, plant nutrient content, and farmers’ gross margins in field trials at two sites, El-Jadida and Ksar-Lkbir, during the 2021/2022 growing season, which experienced contrasting rainfall patterns. A subsequent risk analysis evaluated the agronomic and economic efficiency of these strategies across ten growing seasons (2012–2021), considering variations in fertilizer and wheat market prices. The results indicate that the APSIM-based decision rules, which integrate historical climate data and current in-season rainfall, were the most effective in 2021/2022. This approach optimized farmers’ gross margins in El-Jadida and minimized economic losses in Ksar-Lkbir. The risk frequency analysis highlighted the superiority of APSIM-based decision rules, particularly those incorporating current-year rainfall, during favorable seasons. However, differences between the strategies were less pronounced in other seasons. Overall, the APSIM-based decision tools, which combine historical climate data with current in-season rainfall, and the NPK-engine tool were identified as the most effective strategies for minimizing economic risk in Morocco’s variable and often harsh climate.

Universities and other institutes of higher education could be considered as key actors in the implementation of sustainability pillars, such as the adoption of sustainable practices in wastewater management. However, the adoption of such practices is still an emerging issue. This paper discusses the design and operation of the first combined Oxylag and high rate algal pond (COHRAP) constructed at the university campus in Tunisia for irrigation. Performance was evaluated based on the removal efficiencies of nutrients, chemical oxygen demand (COD), biochemical oxygen demand (BOD), heavy metals, coliforms, and biomass productivity. The potential reuse of sludge and algal biomass is discussed based on the Tunisian national standard regulation for sludge reuse in agriculture (NT 106.20) and the European regulation (EC, 2019/1009) for fertilizer products. Effluent phytotoxicity is tested on the germination and growth on Zea mays L. The results indicate that the COHRAP performance was globally satisfactory; however, biomass productivity (1.4 g m−2d−1) was low, indicating the need for adjustments in the operational parameters. Despite the effluent limitations for TSS and Hg, no phytotoxic effect was observed. Regarding the heavy metal content in sludge and algal biomass, the results obtained were in compliance with NT 106.20 and EC, 2019/1009), respectively. The energy consumption of COHRAP is 1.05 kWh/m3 resulting in operational costs of 0.29 euros/m3. This study revealed that COHRAP could be a sustainable option to treat wastewater from university campuses with resource recovery. Such a choice can be improved by the implementation of an algae recovery step.

In Burkina Faso, small-scale, community-managed irrigation systems play a crucial role in stabilizing agricultural production and improving food security. Over the past three decades, the state has transferred the management of these irrigation systems to local farmer organizations in the hope of improving efficiency and sustainability. This study assesses the long-term performance of six irrigation perimeters Dakiri, Gorgo, Itenga, Mogtedo, Savili, and Wedbila through an in-depth analysis of governance models, infrastructure conditions, and financial sustainability. Performance indicators such as relative water supply (RWS), gross production per unit of irrigation water (PbIr), and water charge recovery rates were used to assess the effectiveness of farmer-led irrigation management. The results reveal persistent governance and financial challenges as well as issues such as water wastage and low yield persisting, despite decades of implementation of farmer-led management. The degradation of irrigation infrastructure, coupled with declining water fee collection rates, threatens the sustainability of these systems. A comparative analysis of international cases suggests that a hybrid governance model, in which the state provides technical and financial support while strengthening accountability mechanisms, could improve the performance of these irrigation systems. This study recommends a shift towards greater state intervention, improved financial mechanisms, and the adoption of digital monitoring tools to ensure a more efficient and sustainable management framework.

The intensification of anthropogenic pressures, particularly those related to agriculture driven by increasing demands for food and cash crops, generates negative environmental externalities. Assessing these externalities is essential to better identify and implement measures that promote the environmental sustainability of rural landscapes. This study aims to develop a multi-criteria assessment method of the negative environmental externalities of rural landscapes in the northern Benin agricultural basin, based on satellite-derived data. Starting from a 12-class land cover map produced through satellite image classification, the evaluation was conducted in three steps. First, the 12 land cover classes were reclassified into Human Disturbance Coefficients (HDCs) via a weighted sum model multi-criteria analysis based on nine criteria related to the negative environmental externalities of anthropogenic activities. Second, the HDC classes were spatially aggregated using a regular grid of 1 km2 landscape cells to produce the Landscape Environmental Sustainability Index (LESI). Finally, various discretization methods were applied to the LESI for cartographic representation, enhancing spatial interpretation. Results indicate that most areas exhibit moderate environmental externalities (HDC and LESI values between 2.5 and 3.5), covering 63–75% (HDC) and 83–94% (LESI) of the respective sites. Areas of low environmental externalities (values between 1.5 and 2.5) account for 20–24% (HDC) and 5–13% (LESI). The LESI, derived from accessible and cost-effective satellite data, offers a scalable, reproducible, and spatially explicit tool for monitoring landscape sustainability. It holds potential for guiding territorial governance and supporting transitions towards more sustainable land management practices. Future improvements may include, among others, refining the evaluation criteria and introducing variable criteria weighting schemes depending on land cover or region.

Le recours aux microalgues cultivées sur des eaux usées urbaines comme intrants durables en agriculture ou pour produire de l’énergie peut offrir des perspectives prometteuses dans une démarche d’économie circulaire. Cependant, l‘efficacité de leur valorisation dépend étroitement de l’espèce algale utilisée et des conditions de culture. Ce travail vise à évaluer le potentiel fertilisant d’une biomasse algale, composée de Scenedesmus sp. et Chlorella sp. issues d’eaux usées urbaines, sur la croissance de Hordeum vulgare L. ainsi que son aptitude à être valorisée comme substrat pour la méthanisation. La caractérisation de la biomasse algale montre des teneurs en métaux lourds conformes à la norme tunisienne (NT 10.44) relative aux amendements mais présentent certaines limites par rapport aux standards européens, en particulier pour le chrome. Cependant, les essais en pots ont montré que l’irrigation avec les extraits d’algues (10, 20, 40 g/L) entraine une amélioration significative des différents paramètres de croissance (poids sec, longueur des racines et des parties aériennes, teneur en chlorophylle) principalement pour la dose de 20 g/L. Le même traitement a aussi provoqué une amélioration des teneurs en N, P et K dans le sol. Le potentiel méthanogène de la biomasse algale révèle une production de 279,6 mL CH4/g MV pour la biomasse non prétraitée, et de 430,08 mL /g MV pour une biomasse ayant subi un prétraitement à basse température. Cette étude confirme que les microalgues cultivées sur des eaux usées peuvent être utilisées en toute sécurité à des fins agricoles et énergétique dans le contexte méditerranéen, sous réserve d’évaluations expérimentales approfondies.

The dual benefit of wastewater and microalgal biomass is a major advantage of high-rate algal ponds, enabling the environmental valorization of these byproducts. This research explored the effect of treated wastewater on the agri-food species Hordeum vulgare (L.) and its associated weed, Emex spinosa (L.) Campd., along with the effects of algal biomass (primarily composed of Closterium, Chlorella, and Scenedesmus spp.) and Diplotaxis harra leaf powder. Initial pot trials applied microalgae and D. harra at 2, 4, and 6 g·kg−1 soil, also confirming that the treated wastewater met reuse standards and did not affect plant growth. The combined treatment at 4 g·kg−1 led to the highest H. vulgare increases in fresh weight (162.71%), root length (73.75%), and shoot length (72.87%), while reducing E. spinosa shoot and root lengths by 30.79% and 52.18%, and fresh weight by 68.24%. Subsequent field experiments using 1.26 t ha−1 of 0.5-cm-applied D. harra and microalgae powders enhanced H. vulgare growth, while reducing the growth of E. spinosa. The reduction in E. spinosa growth was associated with increased electrolyte leakage and malondialdehyde content. These results support the integration of high-rate algal ponds into agriculture, promoting water reuse and reducing reliance on synthetic fertilizers and herbicides in barley production.

Agricultural terraces are a commonly applied soil and water conservation strategy on steep and intensely cultivated hillslopes. Yet agricultural terraces can lead to an increased incidence of landslides. Nevertheless, their effects on hillslope stability remain poorly studied, especially in the tropical Global South, where terraces are increasingly implemented. Here we investigate to what extent the presence of such terraces may increase the incidence of landslides in the densely-populated northwestern Rwanda. For this, we mapped three important landslide events that were triggered by distinct and intense rainfall events in the region and analyzed the relation of these landslides to the presence of terraces and other controlling factors. Based on an inventory of >4,600 mostly shallow landslides in these three events, we show that landslides are about three times more likely to occur on terraced hillslopes as compared to non-terraced hillslopes. However, our results also demonstrate important variability between the events. While the effect was most pronounced for the largest 2020-event, the other two events, 2016 and 2018, showed a less clear or even negative impact of terracing on landslide occurrence. Furthermore, we observed this effect mainly on moderately to highly susceptible hillslopes and less so in areas with a very high landslide susceptibility. Landslides on terraces also tend to be slightly smaller than their counterparts in non-terraced areas. These findings have important implications for both landslide susceptibility assessment and land management.

In the context of climate change, in-season and longer-term yield predictions are needed to anticipate local and regional food crises and propose adaptations to farmers’ practices. Mechanistic models and machine learning are two modelling options to consider from this perspective. In this study, multiple regression (MR) and random forest (RF) models were calibrated for wheat yield prediction in Morocco, using data collected from 125 farmers’ wheat fields. Additionally, MR and RF models were calibrated both with or without remotely sensed leaf area index (LAI), while considering all farmers’ fields, or specifically to agroecological zoning in Morocco. The same farmers’ fields were simulated using a mechanistic model (APSIM-wheat). We compared the predictive performances of the empirical models and APSIM-wheat. Results showed that both MR and RF showed rather good predictive quality (normalized root mean square errors (NRMSEs) below 35 %), but were always outperformed by the APSIM model. Both RF and MR selected remotely sensed LAI at heading, climate variables (maximal temperatures at emergence and tillering), and fertilization practices (amount of nitrogen applied at heading) as major yield predictors. Integration of remotely sensed LAI in the calibration process reduced NRMSE by 4.5 % and 1.8 % on average for MR and RF models, respectively. Calibration of region-specific models did not significantly improve the predictive. These findings lead to the conclusion that mechanistic models are better at capturing the impacts of in-season climate variability and would be preferred to support short-term tactical adjustments to farmers’ practices, while machine learning models are easier to use in the perspective of mid-term regional prediction.

La mise au point d’une méthode automatique d’estimation des surfaces irriguées par les petits exploitants agricoles en Afrique aux abords des cours d’eau, à partir d’outils libres et de données satellitaires gratuites, reste un défi majeur à cause de la diversité des cultures qui y sont pratiquées, de l’étroitesse des parcelles, de la variabilité des cycles culturaux et de la similarité des réflectances des zones irriguées, des zones humides enherbées et de la végétation arborée riparienne. Cet article visait donc à développer une méthode qui permette d’extraire les surfaces agricoles irriguées par les agriculteurs informels le long des berges du fleuve Comoé au cours de la campagne agricole de saison sèche. Pour ce faire, une image composite, obtenue des images de janvier 2019 des satellites Sentinel-1 et 2, combinée à des indices spectraux dérivés et sensibles aux surfaces irriguées (NDVI, MNDWI et NBR2), a fait l’objet d’une classification supervisée à l’aide du classificateur Random Forest sur la plateforme Earth Engine, après une série de masquages automatiques des sols nus, des surfaces des plans d’eau, des infrastructures et de la forêt galerie. Testée sur des données de janvier 2019, pendant laquelle la plupart des agriculteurs ont mis en place leurs cultures, la méthode proposée permet d’estimer efficacement les superficies irriguées. Elle a permis de distinguer les classes ‘zones irriguées’ et ‘zones humides enherbées’ avec une précision globale de 98 %, un coefficient Kappa de 0,91 et des F-scores respectifs de 99 % et 92 %. L’étude a ainsi montré qu’il est possible de développer à moindre coût une méthode automatique et efficace d’évaluation de surfaces irriguées ripariennes à partir de la plateforme Earth Engine.

The implementation of an automatic method for mapping smallholder irrigated plots along river banks in Africa based on remote sensing remains a major challenge because of the diversity of crops grown, the narrowness of the plots, the variability of crop cycles and the similarity of reflectance of irrigated areas, wetlands herbaceous areas and riparian woody vegetation. The purpose of this paper was to design a method enabling an automatic estimate of irrigated croplands on the banks of the Comoe river. To do so, a composite image based on Sentinel-1 and 2 images and spectral indices NDVI, MNDWI, NBR2 were used for supervised classification thanks to Earth Engine platform. A sequence of automatic masks of bare lands, water bodies, infrastructures and the riparian forest was first implemented, before doing a Random Forest classification. Tested with the January 2019 dataset, the proposed method made it possible to distinguish effectively the 'irrigated areas' and 'herbaceous wetlands' classes with an overall accuracy of 98%, a Kappa coefficient of 0.91 and F-scores of 99% for irrigated areas and 92% for herbaceous wetlands. The study showed that it is possible to successfully implement a low-cost and automatic method for monitoring riparian irrigated plots areas along the Comoe river, using the cloud-based Earth Engine platform.

L’évaluation des ressources fourragères est un élément clé de la gouvernance des crises alimentaires du bétail au Burkina Faso. Cette étude visait l’évaluation, pour la première fois, de la possibilité d’estimer les rendements fourragers des pâturages dans les espaces climatiques du Burkina Faso via l’utilisation de modèles statistiques linéaires uni et multivariés construits à partir de données de biomasse végétale fourragère collectées sur le terrain en 2017, 2018 et 2019, de variables satellitaires phénologiques (indice de végétation de la différence normalisée [NDVI] et fraction de rayonnement photosynthétiquement actif absorbé [FAPAR]) et agroclimatiques (précipitations, humidité du sol, évapotranspiration, température de surface). Une recherche exhaustive des meilleurs modèles statistiques linéaires comportant une à quatre variables a été réalisée et les meilleurs modèles selon le critère d’information bayésien (BIC) identifiés. La performance des modèles uni à quadrivariés obtenus s’est avérée assez faible avec, pour l’ensemble des espaces climatiques excepté l’espace sahélien, des RRMSE press variant de 55 % à 61 % (R² press de 0,07 à 0,36), et pour l’espace climatique sahélien des RRMSE press variant de 42 % à 49 % (R² press de 0,59 à 0,69). La baisse de corrélation de la majorité des variables avec la biomasse végétale fourragère selon le gradient nord-sud résulte en une baisse de performance des modèles selon ce gradient. Les variables agroclimatiques se sont révélées inutiles, et celles issues du FAPAR sont globalement plus performantes que celles issues du NDVI. Une très faible plus-value des modèles multivariés comparés aux modèles univariés a été observée, excepté pour l’espace sahélien. Les modèles développés sur des espaces climatiques plus homogènes se sont montrés plus performants. Une série de recommandations a été identifiée pour améliorer le couplage entre données de biomasse végétale fourragère collectées sur le terrain et variables extraites des images satellitaires, et ainsi améliorer la performance des modèles.

In the literature, the studies on wastewater treatment process have shown a great interest in aeration system, most of them were performed on small pilot lab-scale. However, it still an important need to provide useful information aiming at reducing the operating costs and providing decision support to choose aeration devices. Considering that membrane bioreactors operating with high biomass concentrations up to 20-30 g of suspended solids (SS) per liter, the present study was carried out on a larger pilot lab-scale ( +/- 500 l) of insufflation system and mechanical surface aeration on the same fresh sludge with concentration varying between 4.6 and 19.8 g/l. The comparison with literature data shows good correspondence with values obtained with our insufflation system but a great difference from values obtained in a mechanical system. In addition, this research provides a better insight on the calculation method that can be used to estimate the critical parameters of the oxygen transfer, and analyses the feasibility of air diffusion substitution by surface aeration system of Louvain La Neuve wastewater treatment plant of 13000 population equivalent. This extrapolation proved that a saving of +/- 50,000 per year can be achieved by choosing surface aeration rather than aeration by air insufflation.

The main purpose of this study was to analyze observed meteorological data and local people’s perceptions regarding climate change and climate change-related drought in rural areas of Daklak province in the Central Highlands of Vietnam, located in the Lower Mekong Basin. A sample of 354 households was selected using a simple random sampling method. Data were collected from face-to-face interviews with respondents based on a structured questionnaire. Chi-square analysis revealed significant differences in climate change perceptions across the 8 communes studied (χ2 = 23.58, df = 7, p = 0.0013), indicating that location plays a crucial role in shaping these perceptions, primarily based on education and ethnicity. The study found a significant difference in climate change perceptions and observations of climate change-related extreme events, depending on socio-economic and demographic household characteristics. Education, preferred media sources, and income sources positively impacted climate change understanding. The study recommends targeted outreach to less educated communities and those with limited media access. The limits of human perception of slow processes such as climate change or processes noised by strong interannual variability are also shown. The statement highlights none of the discrepancy between farmer’s perceptions and meteorological data relating to clear trends such as global warming. However, when describing trends in precipitation (volume and duration), there is often disagreement between farmer’s perceptions and meteorological observations. This discrepancy can be attributed to less pronounced precipitation trends and high interannual variability. Regarding drought occurrence, most (95

Climate change and rising atmospheric CO2 levels are critical factors influencing agricultural productivity, particularly in Morocco, where cereal crops are essential for food security. The primary objective of this study is to evaluate the combined effects of atmospheric CO2 variations and climatic changes on cereal yields up to 2099 using the CARAIB dynamic vegetation model. This evaluation is driven by four future scenarios based on the Euro-CORDEX initiative's regional climate models under the Representative Concentration Pathway 8.5. As part of this evaluation, the study also validates the CARAIB model for Morocco's major cereal crops: soft wheat, durum wheat, and barley, in order to ensure the model's accuracy in simulating crop responses under projected environmental conditions. The CARAIB model effectively simulated historical cereal yield trajectories across major farming regions in Morocco from 2000 to 2016, demonstrating its robust predictive capability. Our future projections suggest that elevated CO2 levels might initially sustain cereal yields at approximately 90 % of current levels until 2050. This trend indicates that the increase in atmospheric CO2 may exert a moderating influence on the negative impacts of other environmental stressors on crop yields. However, despite this initial buffering effect, the overall yield trend from the present until 2099 indicates a decrease for most combinations of crop, zone, and climate model, even with the CO2 fertilization effect, except in some cases, the model exhibits slight increases or stabilization in yields. Additionally, the CARAIB model predicts potential yield shocks of 10-35 % below current levels from the 2080 s onwards, primarily due to periodic droughts. This variation underscores the complexity of the interplay between CO2 fertilization and climatic changes, emphasizing the urgent need for Morocco to develop adaptive agricultural strategies for long-term sustainability in the face of climatic challenges.

Understanding crop phenology is crucial for predicting crop yields and identifying potential risks to food security. The objective was to investigate the effectiveness of satellite sensor data, compared to field observations and proximal sensing, in detecting crop phenological stages. Time series data from 122 winter wheat, 99 silage maize, and 77 late potato fields were analyzed during 2015–2017. The spectral signals derived from Digital Hemispherical Photographs (DHP), Disaster Monitoring Constellation (DMC), and Sentinel-2 (S2) were crop-specific and sensor-independent. Models fitted to sensor-derived fAPAR (fraction of absorbed photosynthetically active radiation) demonstrated a higher goodness of fit as compared to fCover (fraction of vegetation cover), with the best model fits obtained for maize, followed by wheat and potato. S2-derived fAPAR showed decreasing variability as the growing season progressed. The use of a double sigmoid model fit allowed defining inflection points corresponding to stem elongation (upward sigmoid) and senescence (downward sigmoid), while the upward endpoint corresponded to canopy closure and the maximum values to flowering and fruit development. Furthermore, increasing the frequency of sensor revisits is beneficial for detecting short-duration crop phenological stages. The results have implications for data assimilation to improve crop yield forecasting and agri-environmental modeling.