Landscape　patterns　in　a　region　have　different　sizes,　shapes　and　spatial　arrangements,　which　contribute　to　the　spatial　heterogeneity　of　the　landscape　and　are　linked　to　the　distinct　behavior　of　thermal　environments.　There　is　a　lack　of　research　generating　landscape　metrics　from　discretized　percent　impervious　surface　area　data　(ISA),　which　can　be　used　as　an　indicator　of　urban　spatial　structure　and　level　of　development,　and　quantitatively　characterizing　the　spatial　patterns　of　landscapes　and　land　surface　temperatures　(LST).　In　this　study,　linear　spectral　mixture　analysis　(LSMA)　is　used　to　derive　sub-pixel　ISA.　Continuous　fractional　cover　thresholds　are　used　to　discretize　percent　ISA　into　different　categories　related　to　urban　land　cover　patterns.　Landscape　metrics　are　calculated　based　on　different　ISA　categories　and　used　to　quantify　urban　landscape　patterns　and　LST　configurations.　The　characteristics　of　LST　and　percent　ISA　are　quantified　by　landscape　metrics　such　as　indices　of　patch　density,　aggregation,　connectedness,　shape　and　shape　complexity.　The　urban　thermal　intensity　is　also　analyzed　based　on　percent　ISA.　The　results　indicate　that　landscape　metrics　are　sensitive　to　the　variation　of　pixel　values　of　fractional　ISA,　and　the　integration　of　LST,　LSMA.　Landscape　metrics　provide　a　quantitative　method　for　describing　the　spatial　distribution　and　seasonal　variation　in　urban　thermal　patterns　in　response　to　associated　urban　land　cover　patterns.　(C)　2015　Elsevier　B.V.　All　rights　reserved.