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) for information on this resource. |
| Type: | Resource | |
| Storage: | The size of this resource is 584.6 KB | |
| Created: | Jun 09, 2026 at 5:44 p.m. (UTC) | |
| Last updated: | Jun 11, 2026 at 10:50 p.m. (UTC) (Metadata update) | |
| Published date: | Jun 11, 2026 at 10:50 p.m. (UTC) | |
| DOI: | 10.4211/hs.d314b7e633024ee58649414468ad77f8 | |
| Citation: | See how to cite this resource | |
| Content types: | Single File Content CSV Content |
| Sharing Status: | Published |
|---|---|
| Views: | 45 |
| Downloads: | 1 |
| +1 Votes: | Be the first one to this. |
| Comments: | No comments (yet) |
Abstract
Groundwater forecasts that support sustainable aquifer management often account for climate and hydrologic uncertainty, but they typically assume that human pumping behavior remains stable over time. In intensively irrigated aquifers, this assumption may not hold because pumping decisions can shift with drought, crop choice, energy costs, irrigation technology, regulation, and conservation programs. We examine how non-stationary pumping behavior affects coupled human--groundwater prediction using annual pumping-depth data for 43 county-level agents in the High Plains Aquifer Hydrologic Observatory Area within the Ogallala Aquifer. Using the 1993--2020 pumping record, our workflow identifies where pumping behavior departs from stationarity, localizes when these shifts occur, compares stationary and regime-aware data-driven pumping models, and propagates pumping-prediction uncertainty through the Republican River Compact Administration MODFLOW groundwater model. Results show that non-stationarity occurred in a minority of eight agents and was more clearly detected at the county-agent scale than in aggregated cluster means. Regime-aware modeling better captured post-transition pumping-depth trajectories for seven of the eight non-stationary agents. After propagation through the groundwater model, however, improvements were less consistent: regime-aware simulations better represented groundwater-level trajectories for five agents. The coupled simulations show that uncertainty in changing pumping behavior can widen the range of plausible groundwater outcomes over time. These findings identify behavioral non-stationarity as an important source of groundwater-forecast uncertainty and provide a framework for evaluating when coupled human--water models should update behavioral assumptions and propagate behavioral uncertainty.
Subject Keywords
Coverage
Spatial
Content
readme.md
Data dictionary — HydroShare deposit (Hu, 2026b)
Input data to reproduce Hu & Xi. Place these under data/ in the code
repository. License: CC-BY 4.0. Coverage: 1993–2020, High Plains Aquifer Hydrologic
Observatory Area (Ogallala). Irrigation season months May–October.
Provenance key
- P = project-generated / derived (share here)
- D(src) = derived from third-party source
src(share derived form; cite original) - 3P(src) = essentially third-party
src(consider citing instead of re-hosting)
Files
| File / folder | Contents | Units | Provenance |
|---|---|---|---|
agentdata_{1..48}.csv (43 files) |
Monthly per-agent panel: irrigation depth, crop acreage (corn, wheat, soybeans, sorghum), diesel price, precipitation, temperature, year, month | mm; acres; US$/gal; mm; °C | P (irrigation) + D(GHCNd, USDA, USEIA) (covariates) |
irrigation_depth_monthly_1993_2020.csv |
Monthly irrigation depth by agent/year/month | mm, ft | P (from RRCA pumping) |
irrigation_depth_annual_1993_2020.csv |
Annual pumping volume, irrigated area, depth | acre-ft; acres; ft, mm, in | P |
annual_irrigation_depth.csv |
Wide annual depth, agent × year (1993–2020) | mm | P (aggregate_annual_irrigation.py) |
monthly_crop_data9320.csv (+ .xlsx) |
Monthly commodity prices (corn/soybean/sorghum/wheat) + diesel | US$/bushel; US$/gal | D(USDA, USEIA) |
prcp4rrca9320/monthlyP_*.csv |
Monthly precipitation per RRCA gridcell | mm | D(GHCNd) |
temp4rrca9320/monthlyT_*.csv |
Monthly temperature per RRCA gridcell (Avg/Min/Max) | °C | D(GHCNd) |
agRatio/agAreaR.YYYY (×28) |
Agent → agricultural-area ratio per RRCA gridcell, annual | dimensionless | 3P(RRCA) — not in deposit; cite RRCA |
agRatio/agWatR.YYYY (×28) |
Agent → water amount per gridcell, annual | acre-ft | 3P(RRCA) — not in deposit; cite RRCA |
agRatio/agAreaRM1.YYYY, agAreaRM2.YYYY, agWatRM1.YYYY, agWatRM2.YYYY (×28 each) |
M1 (stationary) / M2 (regime-aware) counterfactual variants for the coupled MODFLOW runs | as above | P |
agRatio/agRatio.csv, agRatioM1.csv, agRatioM2.csv |
Wide annual irrigation-to-precipitation ratio by agent | dimensionless | P |
agRatio provenance. This deposit includes the project-generated
M1/M2variants (agRatio_M1M2.zip) andagRatio*.csv. The baseagAreaR.YYYY/agWatR.YYYYseries are verbatim RRCA-format MODFLOW inputs and are not deposited — obtain them from the RRCA MODFLOW-2000 model (cite McKusick, 2003).
Units summary
irrigation/pumping depth = mm (also ft, in); precipitation = mm; temperature = °C; crop prices = US$/bushel; diesel = US$/gallon; water amount = acre-feet; area/irrigation ratios = dimensionless.
Agent key
- 46 county-level RRCA decision units; 43 with complete 1993–2020 records are used.
- Agent IDs present: {1–32, 36–40, 43–48}; absent: {33, 34, 35, 41, 42}.
- DTC clusters (k=2): Cluster 2 (minority, high-variability) = {2, 3, 24, 28, 29}; Cluster 1 = the other 38.
- Non-stationary agents (BOCPD, p ≥ 0.3) = {2, 12, 14, 18, 20, 24, 28, 29} (8 agents). Agent 3 is in Cluster 2 but did not exceed the threshold and is excluded from predictive analysis.
- Operational changepoints cp*: 2004 (12, 14, 18, 24), 2005 (20), 2011 (29), 2012 (2, 28).
- TODO (Yao): add the agent-ID → county-name lookup if you want county labels public (not currently in the repo).
How to Cite
This resource is shared under the Creative Commons Attribution CC BY.
http://creativecommons.org/licenses/by/4.0/
Comments
There are currently no comments
New Comment