LSMC 2011-2012 - LSMC 2011-2012: Large-scale parallel Monte Carlo simulations for ocean colour applications
Nov 2011 - Oct 2012
This proposal is a follow-up to the previous project on large-scale Monte Carlo (MC) simulations for ocean colour (OC) application studies. The proposed project aims to 1) support OC studies on uncertainty assessments of field above-water radiometry; and 2) explore different parallelization schemes to efficiently handle a large volume of data resulting from threedimensional MC simulations.
Coordinated by: Tamito Kajiyama (CITI), Davide D´Alimonte (CENTRIA), and Jose Cardoso Cunha (CITI).
field above-water radiometry. The underpinning of the proposed studies is the importance of in-situ radiometric measurements in validation and calibration activities for satellite OC observations in current and future space missions. One of the most relevant examples is AERONET-OC, a component for autonomous abovewater measurements operated within the AERONET sensor networks federation managed by NASA Goddard Space Flight Center. AERONET-OC aims to support long-term satellite OC investigations through cross-site consistent and accurate in-situ measurements. The scope of the proposal is to quantify uncertainty budgets due to environmental perturbations (e.g., sun/sky glints induced by sea-surface waves) and measurement protocols (e.g., effects of the integration time required for a single measurement by sensors) in field above-water measurements performed by autonomous radiometer systems in AERONET-OC. This objective requires an extended capability of the MC code to analyze the dependence of uncertainty on the variability of sea-surface geometry not only along the wave propagation direction but also in the direction across it. The proposed project thus intends to enhance the code to perform MC simulations in a threedimensional (3D) domain. Different parallelization and domain decomposition schemes are explored to efficiently handle a large volume of data resulting from 3D MC simulations.
(A total of 40000 CPU-hours and 20 GB disk space were awarded to the present project for consumption of the Milipeia supercomputer at the University of Coimbra.)
