Machine Learning and Spatio Temporal Analysis for Assessing Ecological Impacts of the Billion Tree Afforestation Project - Research article

Aug 3, 2026

This study evaluates the Billion Tree Afforestation Project (BTAP) in Pakistan’s Khyber Pakhtunkhwa (KPK) province by combining remote sensing with machine learning. Random Forest classification on Sentinel-2 imagery shows tree cover rising from 25.02% (2015) to 29.99% (2023) and barren land declining from 20.64% to 16.81%, with accuracy above 85%. Hotspot and clustering analyses indicate vegetation recovery. A predictive NDVI model, supported by SHAP, links soil moisture and precipitation to growth, achieving R2=0.8556 and RMSE=0.0607.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages29

Reading Time1 h 13 min

AccessFree to download

Introduction

Afforestation and environmental relevance

Remote sensing for monitoring afforestation

Machine learning for land-cover classification

Ecology and Evolution RESEARCH ARTICLE OPEN ACCESS Machine Learning and Spatio Temporal Analysis for Assessing Ecological Impacts ofthe Billion Tree Afforestation Project Kaleem Mehmood1,2,3 | Shoaib Ahmad Anees4 | Sultan Muhammad3 | Fahad Shahzad5 | Qijing Liu1,2 | Waseem Razzaq Khan6 | Mansour Shrahili7 | Mohammad Javed Ansari8 | Timothy Dube9 1College of Forestry, Beijing Forestry University, Beijing, China | 2Key Laboratory for Silviculture and Conservation of Ministry of Education, Beijing Forestry University, Beijing, China | 3Institute of Forest Science, University of Swat, Swat, Pakistan | 4Department of Forestry, The University of Agriculture, Dera Ismail Khan, Pakistan | 5Precision Forestry Key Laboratory of Beijing, Beijing Forestry University, Beijing, China | 6Department of Forestry Science and Biodiversity, Faculty of Forestry and Environment, Universiti Putra Malaysia, Serdang, Malaysia | 7Department of Statistics and Operations Research, College of Science, King Saud University, Riyadh, Saudi Arabia | 8Department of Botany, Hindu College Moradabad (Mahatma Jyotiba Phule Rohilkhand University Bareilly), Moradabad, India | 9Institute for Water Studies, Faculty of Science, University of the Western Cape, Cape Town, South Africa Correspondence: Shoaib Ahmad Anees ([anees.shoaib@gmail.com](anees.shoaib@gmail.com)) | Waseem Razzaq Khan ([khanwaseem@upm.edu.my](khanwaseem@upm.edu.my)) Received: 29 September 2024 | Revised: 28 November 2024 | Accepted: 3 December 2024 Funding: This research was funded by the Universiti Putra Malaysia, Malaysia, Vote No. 9750500. Keywords: afforestation | land-use change | machine learning | NDVI | remote sensing | Sentinel-2 ABSTRACT This study evaluates the Billion Tree Afforestation Project (BTAP) in Pakistan's Khyber Pakhtunkhwa (KPK) province using remote sensing and machine learning. Applying Random Forest (RF) classification to Sentinel-2 imagery, we observed an increase in tree cover from 25.02% in 2015 to 29.99% in 2023 and a decrease in barren land from 20. 64% to 16. 81%, with an accuracy above 85%. Hotspot and spatial clustering analyses revealed significant vegetation recovery, with high-confidence hotspots rising from 36.76% to 42.56%. A predictive model for the Normalized Difference Vegetation Index (NDVI), supported by SHAP analysis, identified soil moisture and precipitation as primary drivers of vegetation growth, with the ANN model achieving an R2 of 0.8556 and an RMSE of 0.0607 on the testing dataset. These results demonstrate the effectiveness of integrating machine learning with remote sensing as a framework to support data-driven afforestation efforts and inform sustainable environmental management practices. 1 | Introduction Afforestation is important to global climate change mitigation, land rehabilitation, and biodiversity enhancement strategies. It has recently been announced that the BTAP Project in Pakistanis one of seven ambitious global initiatives and policy tools emphasizing scaling up forest landscape restoration (Kamal, Ali, and Yingjie 2018; Ullah et al. 2020) . This afforestation activity removes large volumes of CO2 from the atmosphere, which is crucial in combating global warming (Haider et al. 2017; Jallatet al. 2021; Khan et al. 2020) . It also enhances ecosystem services related to carbon sequestration and wildlife conservation (Chen and Zhang 2023; Wang et al. 2022; Wang et al. 2024) . Remote sensing technologies have dramatically changed how afforestation programs are monitored and evaluated, predominantly through high-resolution satellite imagery (Shawky | Kaleem Mehmood and Shoaib Ahmad Anees contributed equally to this work. | | --- | | This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
© 2025 The Author(s). Ecology and Evolution published by John Wiley & So

FAQs

How was BTAP evaluated in this study?

BTAP was assessed using remote sensing data and machine learning, applying Random Forest classification to Sentinel-2 imagery and using NDVI-based modeling to quantify vegetation change.

What land-cover changes were observed between 2015 and 2023?

Tree cover increased from 25.02% in 2015 to 29.99% in 2023, while barren land decreased from 20.64% to 16.81%, with classification accuracy above 85%.

Which variables were identified as key drivers of vegetation growth?

SHAP-supported NDVI prediction indicated soil moisture and precipitation as the primary drivers influencing vegetation growth.

Aug 3, 2026

Upload by Theodora