Hello Friends,

In this article we will be taking a closer look at computational modelling of natural disasters. Computational modelling offers an accurate, adjustable and affordable option for forecasting and preventing devastating natural disasters. The ability to gain knowledge and insight about the natural disaster before it occurs can not only help us warn people who would be effected but also taking proactive steps to prevent them from occurring.

Computational modelling is a process which follows set procedure. The computational modelling process is highly interdisciplinary, often transcending the classical fields of physics, biology, and chemistry and fusing them with computer engineering and computer sciences. The framework of the computational model development process is highly structured. First, the type of the model must be decided by the creator(s). There are many types of models ranging from dynamic (time dependent) or steady state and between physical or stochastic. Usually, for the development of models which are able to study natural systems, the model at its core is dynamic and physical. This is because the model must be able to incorporate the influences of time dependent process upon the natural system. Furthermore, the model must be able to stimulate physical processes. Stochastic models are dependent upon the observed past events and probabilities derived from the past data. However, when studying natural systems in the status quo, it becomes important to recognize that past data may in no way resemble future conditions – as a result of climate change and changing environmental conditions.

However, physical modelling has the downfall in terms of complexity. Physical models are often times very complex. The creation of the model algorithms involve extensive mathematical frameworks to describe the processes which are to be modelled. The development of the algorithm can be split into two separate phases: conceptual and mathematical. Before, the formulation of any mathematical framework, the conceptual web of algorithm must be carefully planned. Once the algorithm is planned and each step of the algorithm is fixed, then the mathematical descriptions and expressions associated with each step of the algorithm can be created and assigned. Therefore, the model algorithm can be created. The algorithm can be coded in a computer language and after some finishing touches such as interfacing and visualization, the model is ready for use.

However, how can these models be used to describe natural phenomena, for example like landslides? Well the steps that must be followed are the exact same as above. However, when we deal with natural systems of such high complexity, there can be another step associated in the modelling: calibration. Calibrating the model against past data incorporates the statistics other calibration processes to refine the model values to fit empirical data as closely as possible. Though the model remains physical at its core, the calibration phase simply tailors the model for a specific landscape or use. Hence we have a fully finished model which can be used to model natural phenomena. The model should be designed to run on a simple PC and hence can be a very affordable yet remarkably accurate way to study the natural systems.

The model I have been working on for the past 11 months has been named COMPEL (COMprehensive Physical Environment Influences Landslide) model. The model is a novel comprehensive modelling scheme which forecasts landslide risks and the associated landslide intensity for an area under future changing environment conditions. In the articles to come we will be looking at the exact scheme of COMPEL and how it can help to revolutionize the way to predict, study and ultimately prevent environment influenced landslides.

If you have any questions or comments please feel free to ask or comment below.

Best Regards,

Nitish 

Photo Credit: nrcan.gc.ca