Mosaic　synthetic　aperture　radar　(SAR)　can　achieve　high　resolution　and　wide　swath　(HRWS)　imaging　by　steering　its　antenna　beam　elaborately.　However,　its　unique　time-frequency　diagram　and　radiometric　variation　gives　challenge　to　the　imaging.　The　conventional　methods　cannot　be　applied　to　mosaic　SAR　imaging　directly.　To　address　these　problems,　this　paper　takes　the　imaging　geometry　into　consideration,　utilizing　time　frequency　transform　to　suppress　the　spectrum　ambiguity.　Then,　the　az-imuth　nonlinear　chirp　scaling　(NCS)　algorithm　is　performed　to　compensate　the　variant　modulation　and　the　pattern　weighting　approach　is　used　to　correct　the　radiometric　variance.　Simulation　results　verify　the　effectiveness　of　the　spectrum　analysis　and　the　proposed　imaging　algorithm.　IEEE