Data　aggregation　has　been　emerged　as　a　basic　approach　in　wireless　sensor　networks　(WSNs)　in　order　to　reduce　the　number　of　transmissions　of　sensor　nodes.　In　this　paper,　we　propose　an　energy-efficient　model　based　on　improved　BP　neural　network　by　particle　swarm　optimization　(PSO-BPNN)　in　WSNs.　The　global　optimized　initial　weights　and　threshold　of　BP　network　are　obtained　by　PSO.　And　then　PSO-BPNN　is　deployed　at　both　the　base　station　(BS)　and　the　node　in　WSNs,　helps　to　find　out　potential　laws　according　to　historical　data　sets.　Only　when　the　deviation　between　the　actual　and　the　predicted　value　at　the　node　exceeds　a　certain　threshold,　the　sampling　value　and　new　model　are　sent　to　BS.　The　experiments　on　ocean　surface　temperature　2008　made　a　satisfied　performance.　When　the　error　threshold　greater　than　0.05　degrees　C,　it　can　decrease　more　than　80%　data　transmissions.