Housing Price Prediction/mineral Exploration/natural Disaster Prediction... AI Promotes Innovation in Earth Sciences, Zhejiang University/Tsinghua University/Google Research and Others Have Published Important Results
As a highly interdisciplinary field, Earth science is undergoing a major transformation led by AI. Looking back at 2024, researchers have achieved a series of breakthrough results in smart city construction, housing price forecasting, marine ecological modeling, ground subsidence forecasting, flood forecasting, landslide forecasting, mineral forecasting, etc. These studies not only demonstrate the powerful potential of AI in dealing with complex earth system problems, but also provide innovative solutions for global sustainable development.
This article,HyperAI focuses on AI research in the field of earth sciences. We have selected 15 cutting-edge papers to be interpreted during 2023-2024. Click on the paper title or Chinese interpretation below to jump to the paper interpretation page.Learn more about how AI is driving the future of earth science.
The open source project "awesome-ai4s" brings together more than 100 AI4S paper interpretations and provides massive data sets and tools:
https://github.com/hyperai/awesome-ai4s
Chinese interpretation:Accurately predict Wuhan housing prices! Zhejiang University GIS Laboratory proposes the osp-GNNWR model: accurately describing complex spatial processes and geographical phenomena
Research content:The GIS Key Laboratory of Zhejiang University has improved the accuracy of the model's housing price prediction by introducing an optimized spatial proximity indicator and integrating it into the neural network architecture.
02 、Paper title:OceanGPT: A Large Language Model for Ocean Science Tasks, 2024.05
Chinese interpretation:Selected for ACL 2024! Zhejiang University launches the first ocean language model OceanGPT, making underwater embodied intelligence a reality
Research content:The Zhejiang University team proposed the first large language model in the ocean field, OceanGPT, which can answer questions according to the instructions of oceanographers, demonstrate high professional knowledge in various ocean science tasks, and also acquire preliminary embodied intelligence capabilities in marine engineering.
03 、Paper title:Machine learning-based techniques for land subsidence simulation in an urban area, 2024.02
Chinese interpretation:Beware of urban "chronic diseases": Professor Liu Jianxin's team from Central South University uses AI to predict the risk of land subsidence in the next 40 years
Research content:Professor Liu Jianxin's team from Central South University, in collaboration with the Guangdong Provincial Geological Environment Monitoring Station, the Guangdong Provincial Fourth Geological Brigade, and the University of Boigny in Côte d'Ivoire, established an intelligent prediction model for ground subsidence using extreme gradient boosting regression and long short-term memory networks.
Chinese interpretation:Transfer learning helps! Chengdu University of Technology builds SCDUNet++ model for landslide mapping
Research content:Researchers from Chengdu University of Technology proposed a semantic segmentation model called SCDUNet++, which combines the advantages of convolutional neural networks and Transformer to effectively carry out landslide mapping.
05 、Paper title:Landslide susceptibility modeling by interpretable neural network, 2023.05
Chinese interpretation:A black box becomes transparent: UCLA develops an interpretable neural network (SNN) to predict landslides
Research content:Researchers at the University of California, Los Angeles have developed a superimposed neural network (SNN) that can better analyze the influencing factors in natural disasters and further improve the ability to predict landslide risks.
06 、Paper title:Global prediction of extreme floods in ungauged watersheds, 2024.03
Chinese interpretation:Google's flood prediction model is published in Nature again, beating the world's No.1 system and covering 80+ countries
Research content:The Google Research team has developed a river forecasting model based on machine learning that can reliably predict floods five days in advance. When predicting flood events that occur once every five years, its performance is better than or equivalent to the current prediction of flood events that occur once a year. The system can cover more than 80 countries.
07 、Paper title:Enhanced forecasting of chlorophyll-a concentration in coastal waters through integration of fourier analysis and transformer networks, 2024.09
Chinese interpretation:Deep learning fights against marine red tide crisis! Zhejiang University GIS Laboratory proposes ChloroFormer model to provide early warning of marine algae outbreaks
Research content:Researchers from the GIS Laboratory of Zhejiang University proposed a new deep learning prediction model, ChloroFormer, which can effectively predict the concentration of chlorophyll a in harmful algal blooms in the ocean and provide important information for algal bloom warning.
Chinese interpretation:Better than the five advanced models, the GNNWLR model proposed by Du Zhenhong's team at Zhejiang University: improving the accuracy of mineralization prediction
Research content:A research team from Zhejiang University proposed a new geospatial artificial intelligence method - geographic neural network weighted logistic regression (GNNWLR), which can not only significantly improve the accuracy of mineral predictions, but also improve the interpretability of mineral predictions in complex spatial scenarios.
09 、Paper title:Implicit learning of convective organization explains precipitation stochasticity, 2023.05
Chinese interpretation:Columbia University launches an upgraded version of the neural network Org-NN to accurately predict extreme precipitation
Research content:Columbia University's LEAP Lab used global storm-analyzing simulations and machine learning to create a new algorithm that addresses the problem of missing information and provides a more accurate method for predicting extreme precipitation.
10.Paper title:Deep learning for cross-region streamflow and flood forecasting at a global scale, 2024.05
Chinese interpretation:Analyzing and training data from 2k+ hydrological stations around the world, the Chinese Academy of Sciences team released ED-DLSTM to achieve flood prediction in areas without monitoring data
Research content:A team from the Chengdu Institute of Mountain Hazards and Environment, Chinese Academy of Sciences, proposed a new AI-based runoff and flood prediction model ED-DLSTM to solve the problem of runoff prediction in river basins with and without monitoring data worldwide.
11.Paper title:SuNeRF: Validation of a 3D Global Reconstruction of the Solar Corona Using Simulated EUV Images, 2022.11
Chinese interpretation:AI makes a great contribution! Neural network reconstructs solar images in 3D, revealing the solar poles for the first time
Research content:Researchers at the National Center for Atmospheric Research (NCAR) in Colorado used the NeRFs neural network to convert two-dimensional images of the sun into three-dimensional reconstructions, revealing the sun's poles for the first time.
12.Paper title:Spatial planning of urban communities via deepreinforcementlearning, 2023.09
Chinese interpretation:Defeating 8 human planners: Tsinghua team proposes a reinforcement learning urban spatial planning model
Research content:The research team of Tsinghua University proposed a reinforcement learning model and method for urban community spatial planning, and realized an urban planning process in which human planners collaborate with artificial intelligence algorithms, providing a new idea for the automated planning of smart cities.
13.Paper title: A new paradigm for medium-range severe weather forecasts: probabilistic random forest-based predictions, 2023.02
Chinese interpretation:Colorado State University releases CSU-MLP model to predict medium-term severe weather using random forest algorithm
Research content:Scholars from Colorado State University and SPC jointly released a random forest-based machine learning model, CSU-MLP, which can accurately forecast severe weather in the medium term (4-8 days).
14.Paper title:Social Physics Informed Diffusion Model for Crowd Simulation, 2024.02
Chinese interpretation:Only 5% training samples are needed to achieve optimal performance. Tsinghua University research team released the conditional denoising diffusion model SPDiff to achieve long-range human flow simulation
Research content:A research team from Tsinghua University proposed a conditional denoising diffusion model SPDiff, which can effectively utilize interaction dynamics to simulate crowd behavior through a diffusion process guided by social forces.
15.Paper title:Spatio-Temporal Few-Shot Learning via Diffusive Neural Network Generation, 2024.04
Chinese interpretation:Based on real-life data from seven major cities, the Tsinghua University team open-sourced the GPD model
Research content:The Urban Science and Computation Research Center of the Department of Electronic Engineering at Tsinghua University proposed the GPD model, which uses the diffusion model to generate neural network parameters and transforms spatiotemporal few-shot learning into a pre-training problem for the diffusion model.
