GLAKES-Additional
收藏资源简介:
GLAKES-Additional是由青岛大学开发的全球湖泊数据库,提供了从1990年到2021年全球152,567个湖泊的两年一次的边界划定和面积测量。该数据集通过应用Swin-Unet模型于全球卫星图像,有效地解决了高空间分辨率卫星图像的接收场需求挑战。数据集的创建过程包括使用Swin-Unet模型进行湖泊边界划定和面积提取,以及结合气象数据训练LSTM模型进行湖泊面积变化的预测。该数据集主要应用于湖泊面积变化的监测和预测,旨在通过提高时间分辨率来量化湖泊面积变化,从而更好地理解气候变化和人类活动对湖泊生态系统的影响。
GLAKES-Additional is a global lake database developed by Qingdao University, which provides biennial lake boundary delineation and area measurements for 152,567 lakes worldwide from 1990 to 2021. This database effectively addresses the challenge of reception field requirements for high-spatial-resolution satellite imagery by applying the Swin-Unet model to global satellite images. The development workflow of this dataset includes using the Swin-Unet model for lake boundary delineation and area extraction, as well as training an LSTM model with meteorological data to predict changes in lake area. Primarily applied to the monitoring and prediction of lake area changes, this dataset aims to quantify variations in lake area by improving temporal resolution, thereby facilitating a better understanding of the impacts of climate change and human activities on lake ecosystems.

- 1Constructing a High Temporal Resolution Global Lakes Dataset via Swin-Unet with Applications to Area Prediction青岛大学 · 2024年



