Checking for non-preferred file/folder path names (may take a long time depending on the number of files/folders) ...
This resource contains some files/folders that have non-preferred characters in their name. Show non-conforming files/folders.
This resource contains content types with files that need to be updated to match with metadata changes. Show content type files that need updating.
| Authors: |
|
|
|---|---|---|
| Owners: |
|
This resource does not have an owner who is an active HydroShare user. Contact CUAHSI (help@cuahsi.org) to determine if accessing this resource is possible. |
| Type: | Resource | |
| Storage: | The size of this resource is 36.5 MB | |
| Created: | May 11, 2026 at 1:15 p.m. (UTC) | |
| Last updated: | May 11, 2026 at 2:51 p.m. (UTC) | |
| Citation: | See how to cite this resource |
| Sharing Status: | Discoverable (Accessible via direct link sharing) |
|---|---|
| Views: | 119 |
| Downloads: | 0 |
| +1 Votes: | Be the first one to this. |
| Comments: | No comments (yet) |
Abstract
Trained model checkpoints and reproduction artifacts accompanying the manuscript "Target comparability, not model architecture, limits cross-station river stage prediction" (under review at Water Resources Research, 2026). The bundle contains the LSTM, XGBoost, Manning, Transformer, TCN, and MLP implementations evaluated across 2,885 USGS gauging stations; trained model checkpoints; the frozen canonical 2,020 / 431 / 434 train / val / test station partition; per-station stage normalization statistics; verification scripts (smoke_test.py, compare_to_freeze.py); and a machine-readable record of the paper's headline metrics (frozen_numbers.json). Headline result: re-expressing river stage as low-flow-referenced depth restores cross-station zero-shot generalization (median NSE = 0.708 across 434 unseen test stations), versus systematic failure (NSE = -2.48) when stage is the prediction target. The same pattern holds across six model families.
Subject Keywords
Coverage
Spatial
Content
Credits
Funding Agencies
This resource was created using funding from the following sources:
| Agency Name | Award Title | Award Number |
|---|---|---|
| National Natural Science Foundation of China | Key Program | 52439003 |
| National Key R&D Program of China | None | 2024YFC3211400; 2023YFC3206805 |
| Jiangsu Science and Technology Programme | None | BZ2024034 |
| Major Science and Technology Project of the Ministry of Water Resources | None | SKS-2025092 |
How to Cite
Creative Commons Attribution 4.0 International (CC BY 4.0). Source code (within the bundle) is additionally released under the MIT License upon formal publication of the resource.
https://creativecommons.org/licenses/by/4.0/
Comments
There are currently no comments
New Comment