PTF: an Extensible Component for Sharing and Using Knowledge on Pedo-Transfer functions

Soil data availability for modelling purposes is often insufficient for the application of physical or semi-empirical models simulating soil hydrology. Standard soil surveys frequently do not include hydrological characteristics of the soil, such as either parameters of water retention and conductivity functions or, simpler than the former, estimates of soil water content at field capacity and wilting point. Even when at least part of such data is available, a quality control is needed to ensure not only that values fit within expected ranges, but also to check for consistency across parameters in a specific soil. The use of pedotransfer functions (PTF) allows estimating "what we need from what we have", that is, it allows estimating soil hydrological parameters from soil data often available. The literature makes available a large number of PTF, and new ones are being proposed. Such PTF range from very simple empirical functions, to complex soil physical models. Users must select a PTF to be used based on both available data and their a-priori knowledge about the soil to be simulated. Still, the choice of the PTF to be used is at times controversial, and users may want to compare the estimate made by several PTF against the same data. Also, users may want to test their own PTF, may be specific for a set of soils and thus perfectly adequate for application in a specific contest, against well known ones. An extensible and easily reusable library encapsulating a collection of PTF can be an important tool to support development and operational use of soil-related models, and to share the increasing knowledge about PTF. The objective of this paper is to illustrate the free available component PTF (PedoTransfer Functions). The component is available for both Windows.NET and JAVA platforms, and it is made available with some proof of concept applications (inclusive of source code) in C#, VB.NET and Java, which show how to extend the component and how to use it. The software component presented in this paper meets the following main requirement: i) easy to reuse; ii) with a clear ontology of the variables used in each PTF, where units, value range, and significance, are unambiguously defined; iii) extensible by third parties independently, allowing for an open system to which scientists can contribute; iv) freely available for non-commercial use.