Probabilistic Approach in landscape evolution modeling.

   Landscape evolution can be modeled using either by deterministic or probabilistic approaches (Bates, 2004; Di Baldassarre et al., 2010). In the deterministic approaches calibration of the model uses historical flood or sediment data, use the best-fit model to simulate synthetic design of flood events and sediment yield in 1 in 100 years flood and elaborate the model results in GIS to generate flood hazard maps and erosion maps.  However the prediction of flood and erosion in deterministic approaches use single optimum parameter sets and do not take into account the uncertainties in the modelling processes (Bates et al., 2004) and gives over estimated results.

In the probabilistic approaches, flood-plain mapping generally consists of: construction of flood inundation  and erosion models; sensitivity or uncertainty analysis of the model using the historical flood and sediment data and use of the multiple behavioural which are acceptable models to carry out ensemble simulation using an uncertain synthetic design event as hydrological inputs (Bates et al., 2004). These approaches focus on building an uncertain 1 in 100 year flood extent map. In these approaches, in order to describe the hydraulic behaviour of river and flood-plain dynamics, simplified models are used, as they allow the large number of simulation runs to fully explore the whole parameter space (Di Baldassarre Giuliano et al., 2010).  

For example : CAESAR Lisflood model is a simplified reduced complexity landscape evolution model which follow the probabilistic approach  for simulating long-term flood inundation extent, erosion and deposition map and sediment yield.

Bates, P.D. "Remote Sensing and Flood Inundation Modelling." Hydrological Processes 18 (2004): 2593- 97.

Di Baldassarre, Schumann, Bates, Freer, Beven,.“Flood-plain mapping: a critical discussion of deterministic and probabilistic approaches.” Hydrological Sciences Journal 55(3), (2010): 364- 376.