In order to achieve a carbon-pollution free power sector by 2035, solar and wind energy will need to account for a significant proportion of the electricity supply. Since renewable resources like wind and solar are variable in nature, the accurate and timely prediction of the short-term (minutes-to-days) output from these resources is crucial for the cost-effective and reliable operation of the electric grid.

On May 5-6, 2021, the U.S. Department of Energy (DOE) Solar Energy Technologies Office (SETO) held a virtual workshop on solar forecasting research and development to discuss current state-of-the-art and upcoming advances in topics like data collection, sub-grid modeling, multi-day forecasting, and integration of forecasts into real-world electricity generation. This workshop built on lessons learned from previous SETO Forecasting Workshops in 2016 and 2019.

Keynotes were delivered by Dave Turner from National Oceanic and Atmospheric Administration (NOAA), Justin Sharp from Sharply Focused, and Congcong Wang from Mid-continent Independent System Operator (MISO). A panel of independent system operator (ISO) representatives discussed current capabilities, major challenges, and desired improvements in the practice and application of solar forecasting.

SETO Solar Forecasting II funding program awardees gave presentations covering a range of topics, including improvements in cloud modeling, generation of probabilistic forecasts, use of forecasts for optimal scheduling, and tools for evaluating forecasts. The workshop also included a demonstration of the Solar Forecast Arbiter platform.

Key takeaways, session recordings, presentations, and additional resources can be found below.

The discussions from the breakout sessions provided SETO with the following main takeaways, categorized in four areas: (1) Forecast Models; (2) Solar-related Use Cases; (3) Data and Sensors; (4) Technology Transfer.

Key Areas

Main Takeaways

Forecast Models
  • Support sub-grid scale cloud research
  • Support solar forecasting research and development (R&D) on all timescales
  • Better awareness of National Weather Service/National Oceanic and Atmospheric Administration operational models can lead to better integration of research models
Solar-related Use Cases
  • Support forecasting of behind-the-meter (BTM) solar output
  • Focus on solar-specific impact of extreme weather (including regional conditions such as smoke and snow)
Data and Sensors
  • Need for high-quality and ancillary data
  • Lack of observability of BTM data is still a key issue
  • Assist with access to data (including BTM data)
Technology Transfer
  • Ensure engagement between R&D, vendor, and end-user
  • Solar Forecast Arbiter needs wider outreach
  • Demonstration to operators is key for adoption

SETO thanks all participants for joining this informative and productive workshop. We would like to thank the keynote speakers, panelists, and presenters for providing insights into the latest solar forecasting research advances and the remaining challenges. Finally, we want to thank all participants for their questions and feedback, which will help SETO plan future research topics.

For more information please contact

Session Recordings and Slides

Day 1: May 5, 2021

Day 1 Workshop Recording (password: $&#9*Drs)

Timestamps for different sessions in the recording are included below. For best results, use Google Chrome.


Opening Remarks

Guohui Yuan, DOE Solar Energy Technologies Office

  • Timestamp in recording: 00:03:15

Keynote: The National Oceanic and Atmospheric Administration (NOAA) Atmospheric Science for Renewable Energy Research Program

Dave Turner, NOAA

  • Timestamp in recording: 00:19:15


      Independent System Operator (ISO) Panel

      • Timestamp in recording: 00:51:00

      Forecasting Renewable Resources

      Amber Motley, California ISO

      Solar Forecast at Electric Reliability Council of Texas (ERCOT): Overview and Challenges

      Pengwei Du, ERCOT

      Solar Power Growth in New England

      Mike Fontaine, ISO New England

      New York ISO (NYISO) Solar Forecasting

      Arthur Maniaci, NYISO

      Solar Forecasting in PJM Operations

      Joseph Mulhum, PJM Interconnection

      Elizabeth Anastasio, PJM Interconnection


      Solar Uncertainty Management and Mitigation for Exceptional Reliability in Grid Operations (SUMMER-GO)

      Bri-Mathias Hodge, National Renewable Energy Laboratory (NREL)

      • Timestamp in recording: 01:44:30

      Coordinated Ramping Product and Regulation Reserve Procurements in CAISO and MISO using Multi-Scale Probabilistic Solar Power Forecasts (Pro2R)

      Ben Hobbs, Johns Hopkins University

      Venkat Krishnan, NREL

      • Timestamp in recording: 02:03:40

      Operational Probabilistic Tools for Solar Uncertainty (OPTSUN)

      Aidan Tuohy, Electric Power Research Institute

      • Timestamp in recording: 02:23:40

      Demonstration of Solar Forecast Arbiter

      Will Holmgren, University of Arizona

      • Timestamp in recording: 02:44:20


        Day 2: May 6, 2021

        Day 2 Workshop Recording (password: S76WBKf*)

        Timestamps for different sessions in the recording are included below. For best results, use Google Chrome.


        Keynote: The Importance of Effective Use of Meteorology in the Energy Transition

        Justin Sharp, Sharply Focused

        • Timestamp in recording: 00:01:00

        Keynote: Forecasting in Regional Transmission Organization Reliability Imperative

        Congcong Wang, Midcontinent ISO

        • Timestamp in recording: 00:33:25

        Development of Weather Research and Forecasting (WRF)-Solar v2 – Improving Solar Forecasts

        Larry Berg, Pacific Northwest National Laboratory

        • Timestamp in recording: 01:05:40

        Advancing the WRF-Solar Model to Improve Solar Irradiance Forecast in Cloudy Environments

        Yangang Liu, Brookhaven National Laboratory

        • Timestamp in recording: 01:25:10

        Probabilistic Cloud Optimized Day-Ahead Forecasting System Based on WRF-Solar

        Manajit Sengupta, NREL

        • Timestamp in recording: 01:46:00

        Hybrid Adaptive Input Model Objective Selection (HAIMOS) Ensemble Forecasts for Intra-day and Day-Ahead Global Horizontal Irradiance (GHI), Direct Normal Irradiance (DNI) and Ramps

        Carlos Coimbra, University of California San Diego

        • Timestamp in recording: 02:06:30