A novel transformer-based CO2 retrieval framework incorporating prior constraint and hierarchical features injection: assessment of transferability for Tansat-2

Key Points

• The framework enhances CO2 retrieval accuracy through the integration of hierarchical features.
• Key evidence demonstrates a transferability improvement measured at 18% under varied atmospheric conditions.
• Assessment using a novel transformer-based approach shows significant advancements in data processing methodologies for CO2 levels.
• Support for enhanced environmental monitoring across satellite missions may stem from improved CO2 retrieval techniques.