ProfessorThomas Bishop

In broad terms my research focuses on the use of statistics to model and predict the variation of environmental properties in time and space. In the age of ubiquitous environmental data I believe there is an increasing opportunity for using such models to more precisely manage the environment at finer spatial and temporal scales. Such models are far from developed but I believe some combination of mechanistic and statistical modelling is required to predict in space and time. My main interests are in applying these methods in the domains of agriculture, soil and hydrology.

Currently my main research focus is on developing observation systems for crop yield, soil moisture and soil organic carbon using a combination of mechanistic models, space-time statistics, geospatial data products and field observations. The aim being to nowcast and forecast in near real-time at the spatial resolutions in order of a hectare (100m).

I also have ongoing interests in (i) sampling and monitoring schemes (ii) quantifying the amount and source of uncertainty in models and predictions (iii) change detection in space and time.

I have ongoing funded projects from private and public organisations including:

-using big data to predict cotton yield and constraints to yield;

-development of a decision support tool for irrigation scheduling;

-predict beef quality using metablomics.

In 2010 colleagues and I established a long term catchment-scale monitoring program in Muttama creek catchment in southern NSW where we monitor soil, plant and hydrological properties. It is a validation site for our modelling approaches and the focus of field schools for hydrology and environmental science students.