Building,　a　main　greenhouse　gas　source,　accounts　for　approximately　38%　of　all　carbon　emissions,　and　whole-life-cycle　carbon　emissions　from　building　operations　account　for　60%–80%　of　construction-related　emissions.　However,　the　high-resolution　spatiotemporal　distribution　of　building-operation　carbon　emissions　is　unclear.　We　developed　a　top-down　method　to　calculate　and　map　carbon　emissions　by　(1)　compiling　a　building-operations　carbon-emission　inventory　that　includes　direct　energy　consumption　and　calculating　the　carbon　emissions　of　four　key　energy-intensive　sectors　via　an　emissions-factor　approach;　(2)　cleaning,　reclassifying　and　assigning　weights　to　POI　data　obtained　from　AutoNavi;　and　(3)　using　a　GIS　to　obtain　sectoral　carbon　emissions　in　each　grid　according　to　the　POI　weights　and　to　overlay　and　map　the　total　carbon　emissions.　In　Beijing,　overall　carbon　emissions　peaked　at　95.87　million　tons　(2012).　Emissions　showed　a　clustered　core　with　a　multicenter　discrete　radial　distribution　pattern.　Additionally,　the　carbon　emissions　intensity　peaked　at　487.44　kg/m2　(2015).　The　per-unit-grid　carbon　emissions　distribution　range　was　generally　normally　distributed,　and　the　range　of　103–105　kg/grid　had　the　widest　distribution　and　largest　number　of　associated　grids.　Southeastern　carbon　emissions　hotspots　developed,　and　cold　spots　appeared　along　the　periphery.　This　research　could　help　planners　quickly　zone　and　reduce　carbon　emissions,　leading　to　improved　future　development　strategies.　©　2022　Elsevier　Ltd