Integrated　water　resources　management　from　a　water-energy　nexus　perspective　is　necessary　to　realize　sustainable　development.　However,　the　inherent　uncertainties　and　complexities　in　the　nexus　system　have　posed　great　challenges　for　decision　makers.　In　this　study,　a　novel　two-stage　fuzzy　stochastic　programming　approach　capable　of　addressing　uncertainties　with　both　possibility　and　probability　distribution　is　developed　for　water　resources　management　problems　under　water-energy　nexus.　Under　this　framework,　the　future　uncertain　events　can　be　effectively　defined　as　a　scenario　tree　with　fuzzy　random　information　and　fuzzy　probability　distribution.　Then,　the　proposed　method　is　examined　through　a　real　case　of　Tianjin,　a　coastal　water-stressed　city　in　northeast　China.　A　series　of　decision　alternatives　are　obtained　under　various　water　resource　availability　scenarios　with　the　consideration　of　decision　makers＇　risk　preferences.　Results　showed　that　the　surface　water　and　the　imported　water　accounting　for　70%　of　the　total　water　supply　would　still　be　the　main　contributors　to　support　local　socioeconomic　development　and　that　the　estimated　electricity　consumption　embodied　in　per-unit　water　supply　ranges　from　0.56　kWh/m(3)　to　0.61　kWh/m(3)　through　the　system　nexus　relationship　modeling.　The　results　also　disclosed　that　due　to　climate　change,　water　scarcity　would　require　more　exploitation　of　groundwater　and　utilization　of　desalination　water　and　recycle　water,　which　would　exacerbate　environmental　pressure　and　increase　energy　demand　for　water　supply.　These　findings　can　provide　managerial　insights　and　suggestions　for　decision　makers　to　achieve　flexible　water　management　with　the　consideration　of　system　uncertainties,　water-energy　nexus,　water　shortage,　and　decision　maker＇s　risk　attitude.　(C)　2020　Elsevier　Ltd.　All　rights　reserved.