Applied Chemical Engineering

Deep learning and Multi-Sensor Remote Sensing for predicting Atlas cedar resilience: Integrating Landsat-8, Sentinel-2, and Field Inventories within an AI-Driven Ecological Monitoring Framework

Anass Legdou

Advanced Systems Engineering Laboratory, ENSA Kenitra, Ibn Tofail University, Kenitra, 14000, Morocco

Ayoub Souileh

L3GIE, Mohammadia Engineering School, Mohammed V University in Rabat, Rabat, 10090, Morocco

Achraf Mabrouk

National Forestry School of Engineers, Sale, 11000, Morocco

Said Lahssini

National Forestry School of Engineers, Sale, 11000, Morocco

Bouchra Nassih

Advanced Systems Engineering Laboratory, Faculty of Economics and Management, Ibn Tofail University, Kenitra, 14000, Morocco

Aouatif Amine

Advanced Systems Engineering Laboratory, ENSA Kenitra, Ibn Tofail University, Kenitra, 14000, Morocco

DOI: https://doi.org/10.59429/ace.v9i1.5868

Keywords: Atlas cedar; deep learning; CNN–BiLSTM; Landsat-8 • Sentinel-2; forest resilience; climate change; remote sensing; Morocco; ecological forecasting

---

Abstract

Atlas cedar (Cedrus atlantica) forests in Morocco’s Middle Atlas are experiencing an accelerated decline due to combined climatic and human pressures. Building on previous work on forest transition modeling, this study presents a deep-learning–based framework designed to predict and monitor the ecological resilience of Atlas cedar ecosystems. Multi-sensor satellite images from Landsat-8 and Sentinel-2, combined with field inventory data from the Ain Leuh–Sidi M’Guild massif, were processed to evaluate vegetation health, canopy density, and regeneration potential from 2013 to 2024. A hybrid Convolutional Neural Network–Bidirectional Long Short-Term Memory (CNN–BiLSTM) model was built to capture both spatial and temporal patterns of forest loss and recovery.

Spectral indices such as NDVI, NBR, NDMI, and SAVI were extracted and standardized, while terrain features (altitude, slope, aspect) and bioclimatic variables (temperature seasonality, precipitation during the driest quarter) were included in the model. The hybrid CNN–BiLSTM architecture achieved an overall prediction accuracy of 94.7%, surpassing traditional machine learning methods (Random Forest, SVM, and Gradient Boosting). The spatio-temporal projections reveal a notable decline (−62%) of high-density cedar stands in low-elevation areas, while upper-slope refugia show partial stability and higher regeneration likelihoods.

These results demonstrate the potential of deep learning combined with high-resolution Earth observation data for real-time forest health monitoring and adaptive management. The developed framework provides an operational foundation for Morocco’s Forest Strategy 2020–2030, enabling proactive decision-making for climate-resilient reforestation and ecological restoration in Mediterranean mountain ecosystems.

---

References

[1]. Zaher H, Sabir M, Benjelloun H, Paul-Igor H. Effect of forest land use change on carbohydrates, physical soil quality and carbon stocks in Moroccan cedar area. J. Environ. Manage. 2020;254:109544.

[2]. Camarero JJ, Sánchez-Salguero R, Sangüesa-Barreda G, Lechuga V, Viñegla B, Seco JI, et al. Drought, axe and goats. More variable and synchronized growth forecasts worsening dieback in Moroccan Atlas cedar forests. Sci. Total Environ. 2021;765:142752.

[3]. Linares JC, Pazo Sarria R, Taïqui L, Camarero JJ, Ochoa V, Lechuga V, et al. Efectos de las tendencias climáticas y la degradación del hábitat sobre el decaimiento de los cedrales (Cedrus atlantica) del norte de Marruecos. Ecosistemas 2012;21:7‑14.

[4]. Chafik H, Alaoui SB, Berrada M. Assessing the Dynamics of a South Mediterranean Dryland-Type Forest Kind by Logistic Regression and Cellular Automata [Internet]. In: Chenchouni H, Chaminé HI, Zhang Z, Khelifi N, Ciner A, Ali I, et al., éditeurs. Recent Research on Hydrogeology, Geoecology and Atmospheric Sciences. Cham: Springer Nature Switzerland; 2023 [cité 2025 déc 24]. page 203‑5.Available from: https://link.springer.com/10.1007/978-3-031-43169-2_42

[5]. Beltrame L, Salzinger J, Lampert J, Fanta-Jende P. Towards a Scalable Deep Learning Framework for Forest Monitoring under Challenging Conditions with Multimodal Data [Internet]. 2025 [cité 2025 déc 24];Available from: https://meetingorganizer.copernicus.org/EGU25/EGU25-19171.html

[6]. Sebastiani A, Salvati R, Manes F. Comparing leaf area index estimates in a Mediterranean forest using field measurements, Landsat 8, and Sentinel-2 data. Ecol. Process. 2023;12:28.

[7]. Idbraim S, Bouhsine T, Dahbi MR, Masse A, Arbelo M. Argania Forest Change Detection from Sentinel-2 Satellite Images Using U-Net Architectures [Internet]. In: Kacprzyk J, Ezziyyani M, Balas VE, éditeurs. International Conference on Advanced Intelligent Systems for Sustainable Development. Cham: Springer Nature Switzerland; 2023 [cité 2025 déc 24]. page 174‑84.Available from: https://link.springer.com/10.1007/978-3-031-35248-5_16

[8]. Rhanem M. Aridification du climat régional et remontée de la limite inférieure du cèdre de l’Atlas (Cedrus atlantica Manetti) aux confins de la plaine de Midelt (Maroc). Physio-Géo 2011;143‑65.

[9]. Laaribya S. Analysis of the determinants of the regeneration and growth of Cedar Atlas (Cedrus atlantica (Endl.) Manetti ex Carrière), an endangered endemic taxon in Morocco-Case of the Middle Atlas forests. IOP Conf. Ser. Earth Environ. Sci. 2024;1398:012002.

[10]. Muchsin F, Pradono KA, Prasasti I, Dianovita D, Ulfa K, Veronica KW, et al. EFFECT OF ATMOSPHERIC CORRECTION ALGORITHM ON LANDSAT-8 AND SENTINEL-2 CLASSIFICATION ACCURACY IN PADDY FIELD AREA. Int. J. Remote Sens. Earth Sci. IJReSES 2023;20:57.

[11]. Kim S, Lee Y. Atmospheric Correction of Sentinel-2 Satellite Images for Improvement of Vegetation Indices [Internet]. 2022 [cité 2025 déc 24]. Available from: https://avestia.com/NewTech2022_Proceedings/files/paper/ICEPR/ICEPR_166.pdf

[12]. Navarro-Cerrillo RM, Manzanedo RD, Bohorque J, Sánchez R, Sánchez J, Miguel SD, et al. Structure and spatio-temporal dynamics of cedar forests along a management gradient in the Middle Atlas, Morocco. For. Ecol. Manag. 2013;289:341‑53.

[13]. Legdou A, Amine A, Lahssini S, Chafik H, Berrada M. Predicting Forest Cover Change in Middle Atlas Morocco: A Logistic–CA–Markov Approach [Internet]. In: Bansal JC, Engelbrecht A, Shukla PK, éditeurs. Computer Vision and Robotics. Singapore: Springer Singapore; 2022 [cité 2025 déc 24]. page 229‑39.Available from: https://link.springer.com/10.1007/978-981-16-8225-4_18

[14]. Coudel M, Aubert P, Aderghal M, Hély C. Pastoral and woodcutting activities drive Cedrus atlantica Mediterranean forest structure in the Moroccan Middle Atlas. Ecol. Appl. 2016;26:574‑86.

[15]. Carpentier B, Masse A, Lavergne E, Sannier C. BENCHMARKING OF CONVOLUTIONAL NEURAL NETWORK APPROACHES FOR VEGETATION LAND COVER MAPPING. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2021;XLIII-B2-2021:915‑22.

[16]. Bakhti K, El Amin Arabi M, Chaib S, Djerriri K, Karoui MS, Boumaraf S. Bi-Directional LSTM Model For Classification Of Vegetation From Satellite Time Series [Internet]. In: 2020 Mediterranean and Middle-East Geoscience and Remote Sensing Symposium (M2GARSS). Tunis, Tunisia: IEEE; 2020 [cité 2025 déc 24]. page 160‑3.Available from: https://ieeexplore.ieee.org/document/9105156/

[17]. Kupidura P, Kępa A, Krawczyk P. Comparative analysis of the performance of selected machine learning algorithms depending on the size of the training sample. Rep. Geod. Geoinformatics 2024;118:20240015.

[18]. Stefanus K, Leong H. COMPARISON OF RANDOM FOREST ALGORITHM ACCURACY WITH XGBOOST USING HYPERPARAMETERS. Proxies J. Inform. 2024;7:15‑23.

[19]. Simou MR, Houran N, Benayad M, Maanan M, Loulad S, Rhinane H. MONITORING FOREST DEGRADATION OVER FOUR DECADES USING REMOTE SENSING AND MACHINE LEARNING CLASSIFICATION ALGORITHMS IN BOUSKOURA, MOROCCO. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2024;XLVIII-4/W9-2024:337‑42.

[20]. Barakat A, Brhaiberh M, Ettaqy A, El Jazouli A, Ennaji W. Machine learning based forest species susceptibility mapping using Sentinel 2A data and GIS: A case study of Ait Bouzid forest (Central High Atlas, Morocco). Bull. Geogr. Phys. Geogr. Ser. 2023;63‑78.

[21]. Alibakhshi S. A robust approach and analytical tool for identifying early warning signals of forest mortality. Ecol. Indic. 2023;155:110983.

[22]. Arar A, Nouidjem Y, Bounar R, Tabet S, Kouba Y. Modeling of the current and future potential distribution of Atlas cedar (Cedrus atlantica) forests revealed shifts in the latitudinal, longitudinal and altitudinal range towards more humid conditions. Ecol. Quest. 2020;31:49.

[23]. Linares JC, Taïqui L, Camarero JJ. Increasing Drought Sensitivity and Decline of Atlas Cedar (Cedrus atlantica) in the Moroccan Middle Atlas Forests. Forests 2011;2:777‑96.

[24]. Jurado J, Ramos M, Enríquez C, Feito F. The Impact of Canopy Reflectance on the 3D Structure of Individual Trees in a Mediterranean Forest. Remote Sens. 2020;12:1430.

[25]. Coudel M, Aubert P, Aderghal M, Hély C. Pastoral and woodcutting activities drive Cedrus atlantica Mediterranean forest structure in the Moroccan Middle Atlas. Ecol. Appl. 2016;26:574‑86.

[26]. Coudel M, Aubert P, Aderghal M, Hély C. Pastoral and woodcutting activities drive Cedrus atlantica Mediterranean forest structure in the Moroccan Middle Atlas. Ecol. Appl. 2016;26:574‑86.

[27]. Bonannella C. Spatiotemporal modeling of vegetation dynamics in a changing environment: combining earth observation and machine learning [Internet]. 2024 [cité 2025 déc 24];Available from: https://research.wur.nl/en/publications/bca0991e-20ab-4c38-b91b-6eded1de8b3a

[28]. Williams T, Martinuzzi F, Camps-Valls G, D. Mahecha M. Evaluating Forest Resilience in Europe with Deep Learning Persistence Analysis [Internet]. 2025 [cité 2025 déc 24];Available from: https://meetingorganizer.copernicus.org/EGU24/EGU24-20805.html

[29]. Hammi S, Simonneaux V, Cordier JB, Genin D, Alifriqui M, Montes N, et al. Can traditional forest management buffer forest depletion? Dynamics of Moroccan High Atlas Mountain forests using remote sensing and vegetation analysis. For. Ecol. Manag. 2010;260:1861‑72.

[30]. Linares JC, Pazo Sarria R, Taïqui L, Camarero JJ, Ochoa V, Lechuga V, et al. Efectos de las tendencias climáticas y la degradación del hábitat sobre el decaimiento de los cedrales (Cedrus atlantica) del norte de Marruecos. Ecosistemas 2012;21:7‑14.

[31]. Moukrim S, Lahssini S, Rifai N, Menzou K, Mharzi-Alaoui H, Labbaci A, et al. Modélisation de la distribution potentielle de Cedrus atlantica Manetti au Maroc et impacts du changement climatique. BOIS FORETS Trop. 2020;344:3‑16.

[32]. Seddouki M, Benayad M, Aamir Z, Tahiri M, Maanan M, Rhinane H. USING MACHINE LEARNING COUPLED WITH REMOTE SENSING FOR FOREST FIRE SUSCEPTIBILITY MAPPING. CASE STUDY TETOUAN PROVINCE, NORTHERN MOROCCO. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2023;XLVIII-4/W6-2022:333‑42.

[33]. Dietenberger S, Mueller MM, Adam M, Bachmann F, Stöcker B, Hese S, et al. Digital Forest Inventory Based on UAV Imagery [Internet]. In: IGARSS 2024 - 2024 IEEE International Geoscience and Remote Sensing Symposium. Athens, Greece: IEEE; 2024 [cité 2025 déc 24]. page 4439‑42.Available from: https://ieeexplore.ieee.org/document/10641040/

[34]. Kukunda CB, Beckschäfer P, Magdon P, Schall P, Wirth C, Kleinn C. Scale-guided mapping of forest stand structural heterogeneity from airborne LiDAR. Ecol. Indic. 2019;102:410‑25.