Jobs:Job-00294
From CSDMS
Inverse Modelling, Machine Learning & Uncertainty Analysis
Maynooth University, , Ireland
Apply before: 17 January 2021
Apply before: 17 January 2021
|
Posting:
Position: (postdoctoral) research position
Apply before: 17 January 2021
Apply online:
See attached file:
|
Terrain-AI (T-AI) is a collaborative research project coordinated by Maynooth University, and supported by Science Foundation Ireland’s Strategic Partnership Programme involving Teagasc, TCD, UCD, UL and DCU together with primary Industry partner Microsoft. T-AI’s core R&D activity revolves around improving our knowledge and understanding of Land Use activity - as this relates to Climate Change. A critical component to the success of Terrain-AI is the development and implementation of a suite of model based approaches to improve our understand of the exchanges of energy, water and gases that occur between the land surface and the atmosphere. This exciting role will focus on the development and/or implementation of an inverse modelling based approach (e.g. Stochastic Time‐ Inverted Lagrangian Transport (STILT) model), employing a range of land cover and land use indices, meteorological data fields and other relevant datasets, to exploit the atmospheric measurements of trace gases at Valentia Observatory, Mace Head, Malin head and Carnsore Point. Outputs will be used to evaluate the empirical and process-based model outputs, at landscape scale, and provide a means to constrain model outputs from the wider suite of models being employed. Outputs should also be capable for use in verification of national greenhouse gas inventories. A key challenge for the various models being employed within Terrain-AI will be to bridge the scale gap between plot and landscape while also attempting to quantify the associated uncertainties – recognising that no single ‘optimal’ model or approach exists. This role will undertake an assessment of the uncertainties associated with the various modelling approaches, using a range of techniques (e.g. Bayesian, Machine Learning etc), to develop probabilistic predictions including uncertainty estimates for desired quantities. The candidate will be working closely with PIs, Co-PIs and FIs together with other statistical and computational modelling colleagues at MU as well as collaborating institutions to develop an integrated modelling approach to Land Use Management.
Of interest for: Terrestrial Working Group
