Emissions　from　heavy-duty　diesel　trucks　(HDDTs)　pose　a　major　threat　to　environment　and　human　health.　Understanding　the　behavior　of　pollutant　emissions　from　HDDTs　facilitate　the　formulation　of　traffic-related　policy　measures　to　mitigate　the　adverse　effects.　This　study　proposes　a　new　method　to　estimate　the　emission　inventory　of　HDDTs　and　analyze　their　spatio-temporal　evolution　characteristics.　Multi-source　data　were　fused　to　provide　a　complete　picture　of　the　transport　environment.　With　the　idea　of　modeling　pollutant　emissions　by　＂single　vehicle＂　and　＂road　segment,＂　emission　inventories　were　constructed　with　different　spatiotemporal　scales　using　localized　emission　factors.　A　spatiotemporal　cube　model　was　introduced　to　represent　the　high-resolution　emission　inventory.　A　hot-spot　and　local-outlier　analysis　were　conducted　to　explore　the　spatiotemporal　evolution　mechanism　of　pollutant　emissions.　The　megacity　of　Tianjin　in　China　was　taken　as　the　area　for　the　case　study.　The　average　daily　emissions　of　CO,　NOx,　PM,　and　VOC　are　12,978.18,　48,675.22,　712.6,　and　1217.72　kg　d(-1),　respectively.　Temporally,　the　pollutant　emission　had　a　significant　peak　at　06:00,　11:00,　and　18:00　and　was　affected　by　major　festivals.　Spatially,　the　distribution　pattern　of　emission　was　policy-driven　and　closely　related　to　its　spatial　location.　It　increases　radially　outward　from　the　outer-ring　road　to　the　periphery.　The　hot-spot　analysis　identified　16　pollutant　emission　patterns.　The　road　segment　with　persistent　cold　spots　accounted　for　48.27%　of　the　roads,　mainly　distributed　within　the　outer　ring　road.　The　road　segments　with　intensifying　and　persistent　hot　spots　accounted　for　19.04%　of　roads,　mainly　distributed　on　the　intercity　highway.　The　locations　with　outlier　values　reached　12,027,　accounting　for　31.90%.　The　key　time　intervals　of　the　occurrence　of　the　outlier　pattern　are　11:00-12:00　and　01:00-02:00.　Road　segments　showing　only　the　low-low　cluster　pattern　is　mainly　located　within　the　outer　ring　roads.　Those　with　only　low-high　outlier　pattern　exhibits　a　relatively　scattered　spatial　distribution,　which　is　affected　by　the　heterogeneous　distribution　of　HDDTs.　(C)　2019　Elsevier　Ltd.　All　rights　reserved.