Non-monotonic Benefits of Spatial Detail in Deep Learning for Large-Sample Runoff Prediction

Ouyang, De-Hui

doi:10.5281/zenodo.18606019

Published February 11, 2026 | Version V1

Computational notebook Open

Non-monotonic Benefits of Spatial Detail in Deep Learning for Large-Sample Runoff Prediction

Ouyang, De-Hui (Related person)¹

1. The Hong Kong Polytechnic University

This repository contains the code for the paper "Non-monotonic Benefits of Spatial Detail in Deep Learning for Large-Sample Runoff Prediction". It enables the training and evaluation of four models, including Attr-LSTM, RasterMean-LSTM, MID-CNN-LSTM and HIGH-CNN-LSTM for rainfall-runoff simulations across 531 CAMELS watersheds.

Files

Fig.1.ipynb

Files (1.1 MB)

Name	Size	Download all
Attr-LSTM.py md5:6b3302ef2ac0875d6f69b74582de8774	29.2 kB	Download
Fig.1.ipynb md5:d23235af046730b4bc4bf8c05edaa32b	963.2 kB	Preview Download
Fig.5.ipynb md5:9ad07096794a3fec7b6f5c8929740c23	22.0 kB	Preview Download
Fig.6.ipynb md5:277549515d13bdede445797b0eac54a1	13.3 kB	Preview Download
Fig.7.ipynb md5:f2152bbf51dd896238dd3bdeee78f7a1	21.7 kB	Preview Download
Fig.8.ipynb md5:8b073754d22ecac7ae5a9cad9cb6a1b2	16.2 kB	Preview Download
Fig.9.ipynb md5:37069d2927114ff41728af9a8c4a6bf7	26.0 kB	Preview Download
RasterMean-LSTM, MID-CNN-LSTM and HIGH-CNN-LSTM.py md5:dadcf5551a1d9dfc30f70f0ed5e4753a	52.7 kB	Download

	All versions	This version
Views	12	12
Downloads	0	0
Data volume	0 Bytes	0 Bytes

Non-monotonic Benefits of Spatial Detail in Deep Learning for Large-Sample Runoff Prediction

Authors/Creators

Description

Files

Fig.1.ipynb

Files (1.1 MB)