Peter Joyce et al., (AWAITING PUBLICATION). Contact email: eepjj@leeds.ac.uk This data is provided in order to reproduce findings from a series of experiments with the CASA-TOMCAT model setup. The Carnegie Ames Stanford Approach (CASA) is a land-surface model which was used to produce fluxes of net ecosystem exchange (NEE) and fires. Our simulations held various input parameters constant in CASA (described below). We then forced the TOMCAT atmospheric chemistry model with these data to produce an estimate of atmospheric CO2 at the Barrow observatory in Alaska. Further information on the details of the model setup are described in the 'Model_setup.txt' file. Enclosed in this directory are the simulated CO2 at Barrow observatory, Alaska (71.3N, 156.6E) for a number of experiments which are described below. Half of the text files have variable meteorology and are described as 'atmos_vary' in the title, the remaining half have constant (periodical) meteorology, in which atmospheric transport is retained at 1992 values (and described as 'atmos_const'). Within each file is the time and date of each measurement of atmospheric CO2 in ppm. In order to obtain the required simulated atmospheric value, add the background, ocean and fossil fuel (FF) tracer values to the relevant NEE and fire value from the simulation of interest. The simulations are as follows (with their abbreviation given in parentheses): Constant, periodical temperature scalar (temp) Constant, periodical temperature and moisture scalars (temppre) Constant, periodical fraction of photosynthetically active radiation (fpar) Constant, periodical solar radiation (solrad) All the above variables held constant, periodical (all) Control run in which everything varies (ctrl)