cs.AIMar 29, 2026

TianJi:An autonomous AI meteorologist for discovering physical mechanisms in atmospheric science

Authors: Kaikai Zhang, Xiang Wang, Haoluo Zhao, Nan Chen, Mengyang Yu Jing-Jia Luo, Tao Song, Fan Meng

Organizations: School of Artificial Intelligence, Nanjing University of Information Science and Technology, Nanjing, China · State Key Laboratory of Climate System Prediction and Risk Management (CPRM), Nanjing University of Information Science and Technology, Nanjing, 210044, China · College of Computer Science and Technology, China University of Petroleum, Qingdao, Shandong, China

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Artificial intelligence (AI) has achieved breakthroughs comparable to traditional numerical models in data-driven weather forecasting, yet it remains essentially statistical fitting and struggles to uncover the physical causal mechanisms of the atmosphere. Physics-oriented mechanism research still heavily relies on domain knowledge and cumbersome engineering operations of human scientists, becoming a bottleneck restricting the efficiency of Earth system science exploration. Here, we propose TianJi - the first "AI meteorologist" system capable of autonomously driving complex numerical models to verify physical mechanisms. Powered by a large language model-driven multi-agent architecture, TianJi can autonomously conduct literature research and generate scientific hypotheses. We further decouple scientific research into cognitive planning and engineering execution: the meta-planner interprets hypotheses and devises experimental roadmaps, while a cohort of specialized worker agents collaboratively complete data preparation, model configuration, and multi-dimensional result analysis. In two classic atmospheric dynamic scenarios (squall-line cold pools and typhoon track deflections), TianJi accomplishes expert-level end-to-end experimental operations with zero human intervention, compressing the research cycle to a few hours. It also delivers detailed result analyses and autonomously judges and explains the validity of the hypotheses from outputs. TianJi reveals that the role of AI in Earth system science is transitioning from a "black-box predictor" to an "interpretable scientific collaborator", offering a new paradigm for high-throughput exploration of scientific mechanisms.

Figures & tables

Figure 1: System Architecture Diagram
Figure 2: Workflow Diagram of Hypothesis Generation
Figure 3: Evolution Curve of Hypothesis Quality Scores During Multi-Agent Debate
Figure 4: Spatial distribution of precipitation
Figure 5: Mechanism of Low Soil Moisture Impact on Squall Lines
Figure 6: End-to-End Execution Timeline for Squall Line Experiment
Figure 7: Cumulative System Actions and Fault Self-Recovery
Figure 8: Multi-Agent API Invocation Distribution
Figure 9: Comparison of SST Anomaly Effects on Track and Intensity of Typhoon In-Fa
Figure 10: Nonlinear Mechanism of SST Anomaly Regulating Typhoon Track
Figure 11: Cumulative System Actions and Fault Self-Recovery
Figure 12: End-to-End Execution Timeline for Squall Line Experiment
Figure 13: Multi-Agent API Invocation Distribution
Task NameExecution Result (Summary)
Extract the minimum central pressure every 6 hours, plot the time-pressure evolution line chart, and mark the peak intensity time and value.Successfully extracted data and generated the evolution curve (typhoon_intensity_evolution.png). The analysis shows that the typhoon reached peak intensity at 00:00 UTC on September 14, 2018, with a minimum central pressure of 922.769 hPa.
Track the typhoon movement path and plot the trajectory on a high-resolution map overlaid with the sea level pressure field at peak intensity (12:00 on September 15).Successfully located the typhoon center at each time step and extracted the sea level pressure field at 12:00 on September 15. Generated a combined map of typhoon track and pressure field with high-precision coastlines and latitude–longitude grids (typhoon_track_with_slp.png).
Calculate the total accumulated precipitation field (convective + non-convective), plot the spatial distribution using a meteorological color bar, and mark the maximum precipitation center and value with a red star.Successfully computed the total accumulated precipitation field and generated the spatial distribution map (typhoon_total_precipitation.png). Precisely identified the maximum precipitation center with a total accumulated precipitation of 453.68 mm, highlighted by a red five-pointed star.
Calculate the divergence field based on the 10-m wind field, plot the 2D filled contour map using a divergent color bar, and extract and highlight the extreme strong convergence zone.Successfully computed the horizontal divergence field from U10 and V10 wind components and generated the filled contour map with a divergent color bar (typhoon_divergence_zone.png). Accurately identified the extreme strong convergence region (peak value: -0.00287 s⁻¹), highlighted by a prominent red rectangular box.
Table 1: Summary of Simple Analysis Task Execution Results
Figure 14: Visualization Results of Simple Analysis Tasks
More in this topic

AI Agents for Scientific Discovery

AI agents for scientific discovery use language models, tools, and experiments to generate, test, and refine scientific knowledge.

See all 287 papers

Explore similar work

CardsList
Jun 5, 2026

TianJi-Environ: An Autonomous AI Scientist for Atmospheric Environmental Research

As atmospheric environmental prediction continues to improve, interpretable validation of pollution mechanisms and feedback processes has become a main challenge in atmospheric chemistry. Yet mechanism validation based on complex numerical models still relies heavily on expert knowledge: mechanistic hypotheses must be operationalized into executable experiments, and model outputs must be organized into traceable evidence. We present TianJi-Environ, an auditable AI Scientist for atmospheric-chemistry mechanism validation. TianJi-Environ establishes the first WRF-Chem-based multi-agent framework that autonomously drives complex atmospheric-chemistry simulations, converting mechanistic hypotheses into executable configurations, testing experiments, and evidence criteria. Using ozone response and particulate-matter feedback as two representative examples, we demonstrate TianJi-Environ's capability for mechanism validation. In a summertime ozone case over the North China Plain, the system detects directionally consistent aerosol-radiation-interaction signals in shortwave radiation and boundary-layer height, but judges the evidence for ozone response to NOx control to be incomplete. In a wintertime PM2.5 case over the Guanzhong Basin, it localizes the unsupported link to insufficient propagation from black-carbon perturbation to particulate response and missing diagnostics of vertical absorptive heating. These results show that TianJi-Environ makes expert-driven mechanism validation explicit, structured, and auditable, offering a reproducible paradigm for multi-agent systems coupled with complex atmospheric-chemistry models.

Haoluo Zhao, Hongchun Zhang, Nan Li +6

School of Artificial Intelligence, Nanjing University of Information Science and Technology, Nanjing, China · College of Environmental Science and Engineering, Nanjing University of Information Science and Technology,2026 Nanjing, China · State Key Laboratory of Climate System Prediction and Risk Management (CPRM), Nanjing University of Information Science and Technology, Nanjing, 210044, China +1

Multi-Agent SystemsAutonomous Scientific Discovery
AskPDF
Oct 1, 2026

Strategic Governance of AI Models in Earth Science

AI foundation models pretrained on weather and climate data are increasingly fine-tuned to Earth science tasks well beyond weather forecasting. Their development and adoption are outpacing the scientific community's ability to evaluate them. These models are judged almost entirely by benchmark skill metrics, which measure how closely a forecast reproduces a reference product but not whether a model represents the physical processes governing the system it predicts. Forecast skill and physical reliability are therefore distinct properties. The distinction is most consequential under the nonstationary conditions of a changing climate for which these models were never trained. We identify five priorities for the physical evaluation of AI models in Earth science from task-specific emulators to foundation models, spanning training data, fine-tuning, behavioral testing, mechanistic interpretability, and output validation. We recommend three activities for the coming decade: 1) open AI-ready evaluation datasets, 2) a shared reporting standard for physics-based evaluation, and 3) a dedicated research program on the safety of these models.

Makoto Kelp, Amirhossein Arzani, Patricia Castellanos +6

University of Utah · NASA Goddard Space Flight Center · University of Bristol +5

AI Safety EvaluationClimate ModelingClimate Science
AskPDF
Aug 19, 2026

Tianmu-TC: Physics-constraints Generative Artificial Intelligence for Global Tropical Cyclone Forecasting

Tropical cyclones (TCs) pose severe risks from strong winds and heavy rainfall. However, forecasting their track and intensity remains challenging due to chaotic atmosphere and the rapid amplification of initial condition errors, leading to growing forecast uncertainty. While numerical weather prediction (NWP) and deep learning models have made progress, they remain computationally demanding and often fail under complex meteorological scenarios. Here, we present Tianmu-TC, a physics-constraints generative framework for global TC forecasting. Trained on Western North Pacific data, Tianmu-TC leverages physics-constraints to generate controllable outputs with reduced uncertainty thus improving forecast reliability. Experiments show Tianmu-TC outperforms deterministic and ensemble meteorological artificial intelligence models and authoritative NWP systems such as ECMWF in global ocean basins, with significantly lower computational cost. We further show Tianmu-TC performs well in challenging scenarios such as data sparsity, anomaly tracks, rapid intensification and weakening. These findings suggest physics-constraints generative AI offers a promising approach for reliable, efficient global TC forecasting.

Shiqi Zhang, Pan Mu, Cheng Huang +6

College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou & 310023, China. · School of Control Science and Engineering, Shandong University, Jinan & 250061, China. · School of Computer Science and Engineering, Tianjin University of Technology, Tianjin & 300384, China. +1

Tropical Cyclone Intensity ForecastingWeather ForecastingPhysics-Informed Generative ModelingSpatiotemporal ForecastingProbabilistic Forecasting
AskPDF