My current research focus is at the intersection of meteorology and solar energy

We are in the midst of a massive shift in the way energy is produced and consumed.  At the heart of this change is the rapid uptake of distributed solar energy generation, and on the horizon home energy storage looms.  A future is coming where enough people will be generating solar energy on their rooftops, that large scale changes in cloud cover may destabilise the local electricity grid.  We'll need accurate forecasts of how and when this will occur - and that's where I come in.

Current Projects


Could we use solar panels as cloud sensors?

Could we use solar panels as cloud sensors?

Machine Learning Based Forecasting of Distributed Solar Energy Production

[read the blog post]

  • A pilot project underway in Canberra to trial a solar forecasting network for distributed PV
  • Live data-logging of home solar energy system power output via 3G (100s of systems)
  • Deployment of cloud-imaging devices (10s of locations)
  • Machine Learning and Computer Vision based forecast algorithms
  • Joint project between The ANU and NICTA
 
A negative ramp event in Canberra with a simulation of 12,000+ PV systems

A negative ramp event in Canberra with a simulation of 12,000+ PV systems

Detection and categorisation of significant PV Ramp EVENTS in Australian Capital Cities

  • What weather events cause all of the PV generators in region to suddenly change their power output?
  • Using meteorology, categorise these events so that they can be forecast
  • Using the KPV methodology, produce full scale simulations of tens of thousands of PV generators
  • At proof of concept phase; look for publication in Journal of Applied Meteorology in early 2015

[watch a positive ramp event]

[watch a negative ramp event]


Nerds like me know that model validation is actually fun!

Nerds like me know that model validation is actually fun!

 
We're installing solar radiation measurement and cloud imaging equipment across the ANU campus

We're installing solar radiation measurement and cloud imaging equipment across the ANU campus

Validation and Development of Solar Radiation Models

[read a related blog post]

  • Validation of clear sky radiation models in Australia
  • Validation of diffuse fraction models in southeastern Australia
  • Development of new diffuse fraction models for minute-resolution radiation data

The ANU Solar Radiation and Cloud Measurement Network

[read the publication]

  • Deployment of solar radiation sensors and cloud imaging devices on buildings at The ANU, including:
  • 3x pyrheliometers
  • 7+ pyranometers
  • 3+ sky imagers
  • Monitoring of 6+ PV arrays

Past Research

Air parcel trajectories entering a simulated tornadic vortex

Air parcel trajectories entering a simulated tornadic vortex

Tornadogenesis in Supercell Simulations

  • Work with the Center for Analysis and Prediction of Storms (CAPS)
  • Used the ARPS model to run 50m resolution simulations of supercell thunderstorms
  • These very-high resolution simulations produced tornado-like vortices
  • Used parcel tracing to compute the vorticity budgets of parcels entering the vortices
A Mesoscale Convective System crosses Oklahoma in May 2001

A Mesoscale Convective System crosses Oklahoma in May 2001

Analysis of Mesoscale Convective System (MCS) Storm Outflow

  • Using the Oklahoma Mesonet, completed an analysis of the outflow of thunderstorm systems
  • Particular focus on Mesoscale Convective Systems (MCSs)
  • Analysed each stage in life cycle of the storm separately
  • Results published in Monthly Weather Review
 
/*