Abstract
Abstract. Compositional data, such as soil texture, are hard to deal with in the geosciences as standard statistical methods are often inappropriate to analyse this type of data. Especially in sensitivity analysis, the closed character of the data is often ignored. To that end, we developed a method to assess the local sensitivity of a model output with resect to a compositional model input. We adapted the finite difference technique such that the different parts of the input are perturbed simultaneously while the closed character of the data is preserved. This method was applied to a hydrologic model and the sensitivity of the simulated soil moisture content to local changes in soil texture was assessed. Based on a high number of model runs, in which the soil texture was varied across the entire texture triangle, we identified zones of high sensitivity in the texture triangle. In such zones, the model output uncertainty induced by the discrepancy between the scale of measurement and the scale of model application, is advised to be reduced through additional data collection. Furthermore, the sensitivity analysis provided more insight into the hydrologic model behaviour as it revealed how the model sensitivity is related to the shape of the soil moistureretention curve.
Highlights
In environmental studies, modellers are sometimes confronted with multivariate data that carry only relative information of which the components represent parts of a whole
This is widely known as sensitivity analysis (SA) and allows for (i) the allocation of the uncertainty in the model output to different sources of uncertainty in the model input (Saltelli et al, 2000), (ii) the prioritisation of additional data collection or research concerning the uncertainties identified as most important (Frey and Patil, 2002) and (iii) the verification or validation of a model (Fraedrich and Goldberg, 2000)
5000 textures are sampled from the texture triangle and are used as input to the hydrologic model to evaluate the response of the simulated soil moisture when texture is perturbed with the optimal perturbation factor
Summary
Modellers are sometimes confronted with multivariate data that carry only relative information of which the components represent parts of a whole. A typical example is the sedimentary particle size distribution of which the closed character implies that the components are not free to vary independently such that if one of its components (e.g. clay) decreases (increases), at least one of the others (e.g. silt or sand) must increase (decrease) Because of this particular property, the application of standard statistical methods to compositional data is hampered and many of the results are invalid because the methods are inappropriate to analyse this type of data. A frequently performed statistical exercise involves the evaluation of how changes in the model input or parameters affect the model output This is widely known as sensitivity analysis (SA) and allows for (i) the allocation of the uncertainty in the model output to different sources of uncertainty in the model input (Saltelli et al, 2000), (ii) the prioritisation of additional data collection or research concerning the uncertainties identified as most important (Frey and Patil, 2002) and (iii) the verification or validation of a model (Fraedrich and Goldberg, 2000)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
