@incollection{fdi:010097181,
  title = {{S}hort-term forecasting of land use and land cover changes in {S}enegal's great green wall using {D}eep {L}earning and {S}entinel-2 imagery},
  author = {{D}iouf, {M}.{D}. and {B}a, {M}. and {D}iaw, {S}. and {F}all, {A}hmad and {D}elay, {E}. and {M}asse, {D}ominique and {B}ah, {A}.},
  editor = {},
  language = {{ENG}},
  abstract = {{I}n this paper, we focus on forecasting {L}and {U}se {L}and {C}over ({LULC}) changes in {S}enegal's {G}reat {G}reen {W}all ({GGW}), using {R}emote {S}ensing {I}mages ({RSI}s). {T}he {S}ahel region of {S}enegal faces significant challenges, including land degradation, climate change, drought, and loss of biodiversity. {B}y analyzing historical {LULC} and predicting future developments, this study aims to support the {S}ustainable {D}evelopment {G}oals ({SDG}s) by preserving biodiversity conservation, food security, and resilience to environmental and socioeconomic challenges. {I}n this study, we present {D}eep{C}ascade {GGW}-{F}orecast, a deep learning framework that leverages {S}entinel-2 and corresponding {D}ynamical {W}orld data for 2019, 2021, and 2023 to forecast future trends. {D}eep{C}ascade {GGW}-{F}orecast integrates the {S}at{F}orecaster model to forecast future satellite imagery and {U}-{N}et for semantic segmentation to classify {LULC} classes, including {T}rees, {C}rops, {S}hrub-and-{S}crub and {B}are ground. {S}at{F}orecaster uses {C}onv{LSTM} layers to forecast future satellite imagery achieved a high {R}2 score of 93%. {T}he {U}-{N}et segmentation task, utilizing pretrained encoders ({I}nception-{R}es{N}et-v2 and {R}es{N}et-50), accurately classify {LULC} with 89% {A}ccuracy and 80% {I}o{U}. {T}he {LULC} change analysis quantifies spatiotemporal dynamics, projecting significant increases in {S}hrub-{S}crub, and {T}ree cover by 2025, while {C}rop and {B}are ground are expected to decrease, likely due to ongoing restoration efforts. {T}his insight is crucial for effective environmental management in the region.},
  keywords = {{SENEGAL} ; {SAHEL}},
  booktitle = {{I}ntelligent systems and applications  - {P}roceedings of the 2025 {I}ntelligent {S}ystems {C}onference ({I}ntelli{S}ys)},
  numero = {1660},
  pages = {565--587},
  address = {{C}ham},
  publisher = {{S}pringer},
  series = {{L}ecture {N}otes in {N}etworks and {S}ystems},
  year = {2026},
  DOI = {10.1007/978-3-032-07109-5_39},
  ISBN = {978-3-032-07108-8},
  URL = {https://www.documentation.ird.fr/hor/fdi:010097181},
}