What to Build

Teams are tasked with building an Ethical AI-powered Climate Impact Visualization Platform. The platform should be capable of:

1. Detecting Bias in Citizen-Collected Climate Data: Implementing bias detection algorithms to identify and mitigate potential biases in climate data sourced from TARTLE, a citizen science platform. The platform should utilize statistical analysis, Bayesian models, and multi-method bias correction techniques to ensure the data's reliability and address the inherent challenges of citizen-collected climate records.

2. Generating Explainable AI Predictions: Developing an AI model that makes climate predictions at a local level using bias-corrected data. The AI model should be transparent—using techniques like LIME, SHAP, or Feature Importance Scores—to explain how it arrives at its predictions, ensuring accessibility and clarity for users.

3. Visualizing Localized Climate Impacts: Creating an immersive visualization tool overlaps NOAA's public climate data with personal observations collected through TARTLE. The platform should provide a holistic view of local climate change impacts by combining scientific measurements with individual observations. To enhance user engagement and comprehension, it must feature interactive elements like heat maps, time-series charts, and geospatial mapping.

The resulting platform should adhere to ethical AI principles, focusing on data privacy, informed consent, and transparency. It should allow users to:

• Explore bias-corrected citizen climate data

• View AI-generated local climate predictions

• Interact with visualizations that combine NOAA data and personal observations

• Gain insights into how the AI model makes its predictions

• Understand the local impacts of climate change through various visualization techniques

Key Requirements

• Ethical AI Principles: Prioritize data privacy, informed consent, and transparency in every platform aspect.

• Interactive Features: Include interactive visualizations to help users understand local climate impacts intuitively.

The Final Product Should Allow Users to:

• Explore Bias-Corrected Data: Access and analyze citizen-collected climate data that has been corrected for biases.

• View AI Predictions: Understand local climate predictions generated by explainable AI.

• Engage with Interactive Visuals: Experience visualizations overlaying NOAA data with personal climate observations.

• Gain Insights on AI Models: Understand how the AI generates predictions and the factors influencing outcomes.

• Comprehend Local Climate Impacts: Clearly visualize how climate change affects their local environment.

The Challenge: Bringing Climate Data to Life

Your task is to create a platform that:

1. Bias Detection in Citizen-Collected Climate Data: Implement bias detection algorithms to identify and mitigate potential biases in climate data sourced from the citizen science platform TARTLE. You'll leverage statistical analysis, Bayesian models, and multi-method bias correction techniques to ensure the integrity of the data, addressing potential pitfalls of citizen-collected climate records.

2. Explainable AI Predictions: Develop an AI model that generates local climate predictions from the bias-corrected data. The focus should be on transparency—using tools like LIME, SHAP, or Feature Importance Scores to explain how the AI reaches its conclusions, making the model accessible and understandable to everyone.

3. Localized Climate Change Visualization: Create an immersive visualization tool that combines NOAA's public climate data with personal observations collected through TARTLE. Your tool should give a holistic view of local climate impacts by overlaying scientific measurements with individual stories. Use visual elements like heat maps, time-series charts, and geospatial mapping to engage users.

What to Submit

The Final Product Should Allow Users to:

• Explore Bias-Corrected Data: Access and analyze citizen-collected climate data that has been corrected for biases.

• View AI Predictions: Understand local climate predictions generated by explainable AI.

• Engage with Interactive Visuals: Experience visualizations overlaying NOAA data with personal climate observations.

• Gain Insights on AI Models: Understand how the AI generates predictions and the factors influencing outcomes.

• Comprehend Local Climate Impacts: Clearly visualize how climate change affects their local environment.