Mangroves　are　of　great　significance　to　maintain　the　balance　of　human　production,　life,　and　the　natural　environment　in　coastal　areas,　and　Beibu　Gulf　is　an　important　mangrove　concentrated　area　in　China.　In　this　paper,　the　Beibu　Gulf　of　Guangxi　is　taken　as　the　research　area.　With　the　support　of　Google　Earth　Engine　(GEE)　remote　sensing　big　data　cloud　platform,　Sentinel-2　MSI　and　Sentinel-1　SAR　data　in　2022　are　used　as　data　sources.　According　to　spectral　characteristics,　remote　sensing　index,　phenological　characteristics,　backscattering　characteristics,　and　texture　characteristics,　a　multi-source　data　set　is　constructed.　The　mangrove　remote　sensing　recognition　model　is　constructed　by　using　a　random　forest　algorithm,　and　the　mangrove　extraction　accuracy　of　the　five　models　is　compared　to　determine　the　optimal　model.　The　results　show　that　the　classification　scheme　combining　active　remote　sensing　data,　optical　and　SAR　remote　sensing　indices,　texture,　and　phenological　characteristics　achieves　the　highest　extraction　accuracy,　and　the　overall　accuracy　and　Kappa　coefficient　are　96.23　%　and　0.934,　respectively.　It　shows　that　the　texture　features　and　phenological　characteristics　in　mangrove　distribution　extraction　are　helpful　to　improve　the　accuracy　of　mangrove　extraction.　This　study　verified　the　feasibility　of　remote　sensing　mangrove　extraction　using　phenological　characteristics,　and　the　proposed　method　has　certain　reference　value　for　mangrove　monitoring　in　the　Beibu　Gulf　region.　©　2023　SPIE.