Search for: All records

The authors discuss how exposure to Generative Artificial Intelligence (GenAI) tools led them to reconsider their approach to supporting secondary preservice teachers in curriculum development and to acknowledge both the challenges and opportunities these technologies present in teacher preparation. An upper division undergraduate class designed to support curricular development of all secondary teachers was redesigned to incorporate the use of large language models (LLMs) such as ChatGPT. The findings indicate that GenAI assists in planning, assessment design, and the development of assignments that require human insight beyond AI capabilities. The study highlights the benefits of GenAI in curriculum development and addresses concerns about academic integrity and equitable access. The authors recommend that educators explore GenAI’s potential to support learning and develop strategies for its responsible use in teacher preparation. 

One long-standing technical problem affecting the accuracy of eddy correlation air–sea CO2 flux estimates has been motion contamination of the CO2 mixing-ratio measurement. This sensor-related problem is well known but its source remains unresolved. This report details an attempt to identify and reduce motion-induced error and to improve the infrared gas analyzer (IRGA) design. The key finding is that a large fraction of the motion sensitivity is associated with the detection approach common to most closed- and open-path IRGA employed today for CO2 and H2O measurements. A new prototype sensor was developed to both investigate and remedy the issue. Results in laboratory and deep-water tank tests show marked improvement. The prototype shows a factor of 4–10 reduction in CO2 error under typical at-sea buoy pitch and roll tilts in comparison with an off-the-shelf IRGA system. A similar noise reduction factor of 2–8 is observed in water vapor measurements. The range of platform tilt motion testing also helps to document motion-induced error characteristics of standard analyzers. Study implications are discussed including findings relevant to past field measurements and the promise for improved future flux measurements using similarly modified IRGA on moving ocean observing and aircraft platforms. 

For decades, the use of computer vision as a component of STEM learning has been encouraged at all levels of education—from K-12 to the university levels. A program was developed to support in-service teachers’ development of computer vision. Professional development was provided to middle school teachers while middle school students also attended a summer camp on computer vision. Our research question was: After in-service teachers engaged in artificial intelligence professional development emphasizing computer vision, how did their perceptions of computer vision change? Personal Construct Theory (Kelly, 1955) was used as our methodology. Pairwise comparisons yielded constructs administered in the form of repertory grids. Hierarchical cluster analysis was performed and clusters were identified. Results showed that in-service teachers’ perspectives of computer vision changed after engaging in computer vision-based professional development.