7The investigation of complex geological setting is still dominated by traditional geo-8 data collection and analytical techniques, e.g., stratigraphic logging, dip data 9 measurements, structural ground mapping, seismic interpretation, balance section 10 restoration, forward modelling, etc. Despite the advantages of improving our 11 understanding in structural geometry and fault architecture, the geospatial modelling, 12 applying computer-aided three-dimensional geometric design, visualization and 13 interpretation, has rarely been applied to such complex geological setting. This study 14 used the Lenghu fold-and-thrust belt (in Qaidam basin, NE Tibetan Plateau) to 15 demonstrate that the application of geospatial and geomechanical modelling could 16 improve our understanding and provide an effective technique for investigating the 17 fault architecture and strain distribution. The three-dimensional configuration of the 18Lenghu fold-and-thrust belt was initially derived from traditional analysis techniques, 19 such as regional stratigraphic logging, cross section construction, meso-scale ground 20 mapping and landsat image interpretation. The high-resolution field data and landsat 21 image were integrated to construct the geospatial model, which was subsequently 22 used to quantitatively investigate the fault throw changes along the Lenghu thrust 23 fault zone and to understand its control on the lateral structural variation. The 24 geospatial model was then restored in three dimensions to reveal the kinematic 25 evolution of the Lenghu fold-and-thrust belt. Geomechanical modelling, using a 26 M A N U S C R I P T
A C C E P T E D ACCEPTED MANUSCRIPT2 Mass-Spring algorithm, provided an effective three-dimensional tool for structural 27 strain analysis, which was used to predict the strain distribution throughout the 28 overall structure, e.g., normal faults with throws ranging from meters to tens of 29 meters in the hanging-wall. The strain distribution predicted by geomechanical 30 modelling was then validated by the natural normal faults in the hanging-wall. The 31 high accordance between the strain prediction and statistics of natural normal faults 32 demonstrates good applicability of geospatial and geomechanical modelling in the 33 complex geological setting of the Lenghu fold-and-thrust belt. The geospatial models 34 and geomechanical models, therefore, can provide a robust technique for analyzing 35 and interpreting multi-source data within a three-dimensional environment. We 36 anticipate that the application of three-dimensional geospatial modelling and 37 geomechanical modelling, integrating both multi-source geologic data and three-38 dimensional analytical techniques, can provide an effective workflow for investigating 39 the fault architecture and strain distribution at different scales (e.g., ranging from 40 regional-to meso-scale). 41Keywords 42 geospatial modelling, geomechanical modelling, fault architecture, strain distribution 43 prediction 44