Corrigendum to ’Testing the skill of numerical hydraulic modeling to simulate spatiotemporal flooding patterns in the Logone floodplain, Cameroon’ [Journal of Hydrology, volume 539 (2016) 265-280] (S0022169416302931) (10.1016/j.jhydrol.2016.05.026))

After the publication of this article, we discovered an error in the simulations utilized for the paper. Specifically, a typo in an input file meant that rainfall was not included into the model. Thus, the model simulations in the paper assume no rainfall. We apologize for this involuntary mistake. Nonetheless, after a careful consideration of the study domain and the available hydroclimatic and soil data in the study area, we have concluded that in fact it is reasonable to run the hydraulic model without including rainfall. Existing literature on the Logone Floodplain suggests that much of the early rainfall infiltrates into soils and evaporates quickly. Note that total rainfall over the floodplain is less than one meter per year, whereas annual PET has been estimated as high as three meters per year. Previous studies noted that the soils in the area are clays (this is consistent with the SoilGrids1km described in Hengl et al., 2014) and exhibit surface sealing, but also refer to large cracks in the clays, which would need to fill with water prior to sealing (Naah, 1990). Infiltration and subsequent evaporation from bare soils is not handled in the LISFLOOD-FP hydraulic model that we used. To the extent that this process is important, we would argue that it is better to simply leave precipitation out of the hydraulic simulations. Consistent with previous studies (e.g. Evans and Loth, 2004) we believe that the main function of rainfall over the floodplain is to allow saturation of the soil a few weeks before the overbank flow, thus facilitating, but not controlling, the water movement on the floodplain. Accounting for this effect in the LISFLOOD-FP model requires knowledge of infiltration rates at high spatial resolution, which is currently infeasible in this region of Cameroon. However, given the hydroclimatic and surface characteristics of the study area, we believe that neglecting the effect of rainfall, evaporation, and infiltration will not have a significant effect on the accuracy of simulated flooding dynamics at the spatial resolutions utilized in the paper. Therefore, the error in the simulation set-up does not affect the main conclusions of our work. The following sections of the paper are impacted by this corrigendum: 1. The inclusion of TRMM time series as input into the model as described in the methods section. The model did not include rainfall.2. The interpretation of the figure describing the mass balance of the floodplain (Fig. 6 in the original paper). Precipitation and evaporation data utilized in this figure were taken from a model output representing P-E. The correct interpretation is that precipitation does not compensate evaporation, thereby making the overbank flow the principal factor controlling flooding dynamics (consistent with Evans and Loth, 2004). See Fig. 1, below.3. All the sentences in section 4.3 that show a detailed analysis of TRMM data to explain spatial discrepancies between model output and Landsat-derived flooding maps.4. The sentences in the discussion on the apparent similar importance of rainfall and overbank flow.DOI of original article: 10.1016/j.jhydrol.2016.05.026