Drainage reorganisation and species evolution: model sensitivity analysis data

Data description:

‘trial_factor_values.csv’: The factor values for experiment trials were generated using a quasi-random Sobol sequence (Sobol, 1967). The table field, ‘initial_landscape_id’ is the identifier for unique combinations of the following factor values that controlled the landscape elevation in the initial conditions phase of the model: initial elevation seed, \(U\), \(K\), and \(k_d\). The factors, \(U\), \(K\), \(k_d\), \(P_m\), and allopatric wait time varied logarithmically. The values of these factors in the file are the exponent of base 10.

‘trial_response_values_initial_conditions_phase.csv’: Topographic relief at steady state along with the model time to initial steady state are the trial model responses included in the file. Values are listed for each initial landscape ID rather than trial because many trials had the same combinations of the factors that controlled the topography of the initial landscape. 

‘trial_response_values_perturb_phase_base_level_fall_scenario.csv’ and ‘trial_response_values_perturb_phase_fault_throw_scenario.csv’: Model responses of the perturb phase for base level fall and fault throw scenario along with the initial landscape ID, species count values, and the model time back to steady state.

The files beginning with `sobol`: the sensitivity analysis results output by the software, ‘SALib’ (Herman and Usher, 2017). ‘S1’, ‘S2’, and ‘ST’ in the file name indicates if the file contains data of the Sobol first, second, or total order effect, respectively.