Large Language Models (LLMs) such as ChatGPT possess advanced capabilities in understanding and generating text. These capabilities enable ChatGPT to create text based on specific instructions, which can serve as augmented data for text classification tasks. Previous studies have approached data augmentation (DA) by either rewriting the existing dataset with ChatGPT or generating entirely new data from scratch. However, it is unclear which method is better without comparing their effectiveness. This study investigates the application of both methods to two datasets: a general-topic dataset (Reuters news data) and a domain-specific dataset (Mitigation dataset). Our findings indicate that: 1. ChatGPT generated new data consistently enhanced model’s classification results for both datasets. 2. Generating new data generally outperforms rewriting existing data, though crafting the prompts carefully is crucial to extract the most valuable information from ChatGPT, particularly for domain-specific data. 3. The augmentation data size affects the effectiveness of DA; however, we observed a plateau after incorporating 10 samples. 4. Combining the rewritten sample with new generated sample can potentially further improve the model’s performance.

The SPRUCE experiment is the primary component of the Terrestrial Ecosystem Science Scientific Focus Area at ORNL focused on terrestrial ecosystems and the mechanisms that underlie their responses to environmental change. As of December 2025, the SPRUCE experiment is ending its planned decadal timeframe. The manipulation evaluated the response of the existing biological communities to a range of warming levels from ambient to +9°C, provided via large, modified open-top chambers. The ambient and +9°C warming treatments were also conducted at eCO2 (in the range of 800 to 900 ppm). Both direct and indirect effects of these experimental perturbations have been analyzed to develop and refine models needed for full Earth system analyses. The instruments and infrastructure needed for these measurements include meteorological tower-based CO2/H2O sampling and analysis systems, air temperature and relative humidity probes, precipitation gauges, wind speed and direction instruments, solar radiation sensors, subsurface soil moisture and temperature sensors, water level and conductivity sensors, continuous vegetation sap flow and dendrometer sensors, and companion dataloggers to record and report the data. The data collection system deployed of approximately 80 dataloggers, 1100 sensors, and associated instruments. Through the 10 years of observations, we collected substantial amount of data and want to present variety of data product available for analysis and scientific investigation.

Operational constraints governing hydropower facilities are buried in lengthy regulatory documents with inconsistent formatting, yet no scalable method exists to extract them. Here, we present a regulatory interpretation framework that pairs domain knowledge guided prompt engineering with multi-criteria evidence ranking system, implemented using openly available large language models (LLMs) deployable on local infrastructure. The framework utilizes expert regulatory reasoning, distinguishing mandated minimums from operational capabilities, resolving conflicting values across document sections, and producing traceable, audit-ready outputs (extracted value, regulatory context, and verbatim evidence) and applies it to U.S. Federal Energy Regulatory Commission (FERC) licenses and other hydropower regulatory documents. We evaluated four open LLMs, Llama 3.2 3B, Llama 3 8B, GPT-OSS 20B, and Llama 3.3 70B using a representative set of 50 hydropower regulatory documents across ten operational variables. For simple structured variables (project names, locations, generation capacity), the 70B model achieved 90-98% accuracy, while complex variables requiring interpretation of scattered regulatory text showed substantially lower performance (0-43% for zero-shot extraction). For a complex variable tested (minimum flow requirements), a targeted multi-stage extraction approach improved accuracy from 43% (zero-shot) to 76% (targeted). Overall accuracy across all ten variables was 52.7% (Correct + Partially Correct), with substantial variation by variable complexity. Among the four models tested, only the 70B model was able to follow the complex regulatory reasoning prompts; smaller models failed systematically via false negatives and hallucinations rather than reduced accuracy. The framework shifts regulatory data extraction from document-level classification to fact-level extraction with audit trail, substantially reducing expert effort. The approach is transferable to other regulatory domains where critical information is dispersed across unstructured documents.

Operational constraints governing hydropower facilities are buried in lengthy regulatory documents with inconsistent formatting, yet no scalable method exists to extract them. Here, we present a regulatory interpretation framework that pairs domain knowledge guided prompt engineering with multi-criteria evidence ranking system, implemented using openly available large language models (LLMs) deployable on local infrastructure. The framework utilizes expert regulatory reasoning, distinguishing mandated minimums from operational capabilities, resolving conflicting values across document sections, and producing traceable, audit-ready outputs (extracted value, regulatory context, and verbatim evidence) and applies it to U.S. Federal Energy Regulatory Commission (FERC) licenses and other hydropower regulatory documents. We evaluated four open LLMs, Llama 3.2 3B, Llama 3 8B, GPT-OSS 20B, and Llama 3.3 70B using a representative set of 50 hydropower regulatory documents across ten operational variables. For simple structured variables (project names, locations, generation capacity), the 70B model achieved 90-98% accuracy, while complex variables requiring interpretation of scattered regulatory text showed substantially lower performance (0-43% for zero-shot extraction). For a complex variable tested (minimum flow requirements), a targeted multi-stage extraction approach improved accuracy from 43% (zero-shot) to 76% (targeted). Overall accuracy across all ten variables was 52.7% (Correct + Partially Correct), with substantial variation by variable complexity. Among the four models tested, only the 70B model was able to follow the complex regulatory reasoning prompts; smaller models failed systematically via false negatives and hallucinations rather than reduced accuracy. The framework shifts regulatory data extraction from document-level classification to fact-level extraction with audit trail, substantially reducing expert effort.