Title: ACRIM total solar irradiance satellite composite validation versus TSI proxy models

Abstract: The satellite total solar irradiance (TSI) database provides a valuable record for investigating models of solar variation used to interpret climate changes. The 35-year ACRIM TSI satellite composite was updated using corrections to ACRIMSAT/ACRIM3 results derived from recent testing at the Laboratory for Atmospheric and Space Physics/Total solar irradiance Radiometer Facility (LASP/TRF). The corrections lower the ACRIM3 scale by ~5000 ppm, in close agreement with the scale of SORCE/TIM results (solar constant ~1361 W/m^2). Relative variations and trends are not changed. Differences between the ACRIM and PMOD TSI composites, e.g. the decadal trending during solar cycles 21-22, are tested against a set of solar proxy models, including analysis of Nimbus7/ERB and ERBS/ERBE results available to bridge the ACRIM Gap (1989-1992). Our findings confirm: (1) The validity of the TSI peak in the originally published ERB results in early 1979 during solar cycle 21; (2) The correctness of originally published ACRIM1 results during the SMM spin mode (1981-1984); (3) The upward trend of originally published ERB results during the ACRIM Gap; (4) The occurrence of a significant upward TSI trend between the minima of solar cycles 21 and 22 and (5) a decreasing trend during solar cycles 22-23. Our findings do not support: (1) The downward corrections to originally published ERB and ACRIM1 results during solar cycle 21; (2) A step function sensitivity change in ERB results at the end-of-September 1989; (3) the validity of ERBE's downward trend during the ACRIM Gap or (4) the use of ERBE results to bridge the ACRIM Gap. Our analysis provides a first order validation of the ACRIM TSI composite approach and its 0.037%/decade upward trend during solar cycles 21-22. Thus, solar forcing of climate change may be a significantly larger factor than represented in the CMIP5 general circulation climate models.