In this paper, we present a case study of fracture characterization by integrating borehole data with a variety of seismic attributes in a carbonate reservoir from a giant offshore field, United Arab Emirates. The objectives are to determine to what extent seismic data may be confidently used for mapping spatial distributions of subtle faults and fracture corridors in the reservoirs and to better understand the distribution of overburden anomalies (karsts, high impedance channels) for field development planning. Borehole data used in our study include information from core descriptions (fracture density and orientations), image logs, cross-dipole shear-wave anisotropy analysis, and dynamic data (well testing, PLT, tracer, and mud-loss). The seismic attributes include standard and advanced post-stack geometrical attributes; pre-stack seismic azimuthal AVO attributes, and recently developed pre-stack diffraction imaging. We find that there are common features that can be identified in different attributes, and the differences may indicate different scales of fractures. We also observe a qualitative correlation in the area of history match challenges and high anisotropy magnitude, where seismic anisotropy can identify relatively high fracture intensity regions/zones instead of pinpointing individual fractures and complements other attributes as differences do exist between seismically identified fracture zones and well data due to overburden anisotropy, resolution and sampling issues (which are addressed using the synthetic modeling approach). Diffraction attributes have revealed more detailed geological features in overburden (e.g. karsts) and reservoirs (e.g. lineaments) than in reflection data and a comparison with mud loss data in the shallow zones looks promising with a good correlation between mud loss and collapsed features. This work has provided an improved understanding of the applicability of the using multi-seismic attributes for fracture characterizations in carbonate reservoirs.