MISO Forecasting Challenge: Probabilistic Temperature Forecasting Using Available Weather Data Models

13 Submissions
$15,000 USD
Challenge under evaluation
  1. 1st August 2023

    Submission

  2. 18th September 2023

    Under evaluation

  3. 18th October 2023

    Verification in process

  4. 2nd December 2023

    Awarded

Challenge overview

OVERVIEW

Midcontinent Independent System Operator (MISO) is an independent, not-for-profit organization that delivers safe, cost-effective electric power across 15 U.S. states and the Canadian province of Manitoba. One vital aspect of MISO’s operations is the prediction of electricity generation and demand throughout its service region and this, in turn, depends on having accurate weather forecasts for factors such as temperature, precipitation, wind speed, etc. Currently, MISO utilizes a proprietary ensemble forecast from a supplier that provides deterministic forecasts for near-term (7-10 day) conditions. In this Challenge, MISO is seeking ideas and advanced frameworks that leverage multiple weather forecast data sets to generate probabilistic forecasts specifically tailored to MISO's energy grid operations.

This is a Prize Challenge which requires a written proposal to be submitted. There will be a guaranteed award for at least one submitted solution. To be awarded, Solvers are required to transfer non-exclusive rights to the Intellectual Property (IP) in their proposed solution.

Submissions to this Challenge must be received by 11:59 PM (US Eastern Time) on September 18, 2023.
Late submissions will not be considered.

- Login or register your interest to start solving!

 

ABOUT THE SEEKER & ELIGIBILITY

Midcontinent Independent System Operator (MISO) is an independent, not-for-profit organization that delivers safe, cost-effective electric power across 15 U.S. states and the Canadian province of Manitoba. Operating one of the world’s largest energy markets (with annual energy transactions worth more than $40 billion), MISO is committed to collaborating with all stakeholders to create cost-effective and innovative solutions for our changing industry. To learn more about MISO and its mission, visit www.misoenergy.org  and read about its Strategy and Value Proposition.

Find out more about participation in Wazoku Crowd Challenges.

 

THE CHALLENGE

Background

Midcontinent Independent System Operator (MISO) is an independent, not-for-profit organization that delivers safe, cost-effective electric power across 15 U.S. states and the Canadian province of Manitoba. MISO is committed to reliable, non-discriminatory operation of the bulk power transmission system and collaborating with all stakeholders to create cost-effective and innovative solutions for our changing industry. MISO operates one of the world’s largest energy markets with more than $40 billion in annual gross market energy transactions.   

The evolution of the U.S. electric grid reflects significant changes across load, generation, transmission, distribution, resource availability and need.  MISO must consider the dramatic transformation of the resource mix in MISO’s footprint over this period and the likelihood that similar changes in the energy mix will continue into the foreseeable future. The exploration of these changes has led MISO to focus on three over-arching trends that are transforming and will continue to transform the industry and the MISO region, referred to as the “3Ds”:   

  1. De-Marginalization – refers to the modified set of resources that can provide the next needed, or “marginal,” increment of energy at zero additional costs (e.g., renewables), or very low additional costs (e.g., highly efficient gas-fired generation). 
      
  2. Decentralization – involves the shift away from large, central-station power plants to smaller, often variable resources that are located on local, low-voltage electricity distribution networks, or “behind the meter” at homes and businesses. 
      
  3. Digitalization – refers to the revolution in information and communication technologies and platforms that will continue to disrupt nearly everything in our economy, including energy services.  

Addressing these needs will take time and internal functional coordination from system planning through grid and market operations. Most importantly, it will require a high level of stakeholder collaboration, new partnerships with advanced technology providers, and new coordination with distribution operators to innovate holistic solutions to ensure reliability and create value for end-use consumers.  

As MISO navigates the transition just described, the capabilities and services provided to members and stakeholders will need to evolve as well. This is true of the future state of the grid as well as the transitory period in getting there. MISO will need to develop new capabilities as well as become more effective at our current capabilities. One capability MISO will need to improve upon our ability to synthesize weather forecast data. Weather has a primary effect on both electricity demand and generation and as such has a significant impact on MISO’s operations.

Currently, MISO operates the grid based on a deterministic forecast built on forecast models based on a proprietary combination of the following:

Model Source Link
Global Forecast System (GFS) NOAA https://www.ncei.noaa.gov/products/weather-climate-models/global-forecast

Global Ensemble Forecast System (GEFS)

NOAA

https://www.ncei.noaa.gov/products/weather-climate-models/global-ensemble-forecast

European Centre for Medium-Range Weather Forecasts (ECMWF)

ECMWF

https://www.ecmwf.int/en/forecasts

European Centre for Medium-Range Weather Forecasts Ensemble Prediction System (ECMWF-EPS)

ECMWF

https://www.ecmwf.int/en/forecasts

North American MESOscale Forecast System (NAM)

NOAA

https://www.ncei.noaa.gov/products/weather-climate-models/north-american-mesoscale

Rapid Refresh (RAP)

NOAA

https://rapidrefresh.noaa.gov/

High-Resolution Rapid Refresh (HRRR)

NOAA

https://rapidrefresh.noaa.gov/hrrr/

The current forecasting methodologies predominantly rely on deterministic forecasts, which provide limited insight into the range of potential weather scenarios and their associated probabilities. The challenge is to develop an advanced framework that leverages multiple weather forecast data sets to generate probabilistic forecasts specifically tailored to MISO's energy grid operations. Incorporating probabilistic forecasting techniques into MISO's operational decision-making processes can help improve grid reliability, enhance load management, and enable more effective utilization of renewable energy resources.

 

SOLUTION REQUIREMENTS & ACCEPTANCE CRITERIA

MISO is seeking new ensemble models that leverage multiple weather forecast data sets to generate probabilistic forecasts with an initial focus on temperature. Models may utilize any of the forecast models listed above as well as any additional forecasts Solvers may discover on their own.

MISO is primarily interested in solutions that meet the following requirements:

Must have:

  1. Provide an accurate probabilistic forecast of maximum and minimum daily temperatures for specific locations 7-10 days in advance.
     
  2. Must use a temperature grid of 1 °F for both maximum and minimum temperature forecast and provide the probability that the maximum or minimum temperature for a specific day falls within each one-degree bin.
     
  3. Scalable and extendable to any location within the region served by MISO.
     
  4. Model must be implemented in Python and must be automated in its use. Solvers with top solutions based on a preliminary evaluation of accuracy will be asked to submit source code and documentation of the model implementation for further testing and verification by MISO.

Evaluation

For preliminary evaluation of the accuracy of the proposed model, Solvers are asked to provide hindcast predictions of maximum and minimum temperatures as described above for the locations Indianapolis (KIND), Little Rock (KLIT), and Minneapolis (KMSP) for two historic windows:

  • Winter - Dec 17th through Dec 26th, 2022
  • Summer - June 22nd through July 2nd, 2023

Hindcasts must only utilize weather forecasts available 7 days or more prior to the date period being forecast.

Accuracy of the submitted hindcasts will be evaluated using Continuous Ranked Probability Score (CRPS) and Brier Score. Overall evaluation of solutions will incorporate accuracy as well as other factors such as usability and scientific underpinnings of the methodology.

Solutions with Technology Readiness Levels (TRLs) 1-3 are invited.

 

This is a Prize Challenge, which has the following features: There will be a guaranteed award for at least one submitted solution.

  1. The best solution(s) has the opportunity to win all or part of the guaranteed award of $15,000 USD.
  2. Awards will be contingent upon the theoretical evaluation and experimental validation of the proposal by MISO.

To be awarded, Solvers are required to transfer non-exclusive rights to the Intellectual Property (IP) in their proposed solution.

 

YOUR SUBMISSION

The submitted proposals must be written in English and should include:

  1. An Abstract and optional Conclusion of the proposed solution.
  2. A detailed description of the proposed solution that meets the above Solution Requirements.
  3. A rationale as to why the Solver believes that the proposed solution will achieve the objectives of the Challenge. This rationale should address each of the requirements as described in the Challenge.
  4. Hindcast data as described above.
  5. Any appropriate supporting data, drawings, etc.
  6. MISO may wish to partner with the Solver at the conclusion of the Challenge. Please describe your expertise and include a statement indicating your interest in this opportunity.

Find out more about participation in Wazoku Crowd Challenges.

Submissions to this Challenge must be received by 11:59 PM (US Eastern Time) on September 18, 2023.
Late submissions will not be considered.

After the Challenge submission due date, MISO will complete the review process and make a decision with regards to the winning solution(s) according to the timeline in the Challenge header. All Solvers who submit a proposal will be notified about the status of their submissions.

Use the slider to explore how the Challenge process works:

Register

Review & Accept

Submit

Win

To start solving this Challenge, log in to the Challenge Center or register as a Solver

Visit the Wazoku Crowd Community to start the conversation