Water temperature's cumulative effect induced by the construction of cascade reservoirs in the lower reaches of Jinsha River

Hu Guangyang, Yang Zhonghua, Fengpeng Bai, Ren Yufeng, Yiming Ma, Wenhui LiJournal of Lake Sciences 2025 · DOI

TL;DR

This study quantifies how the massive cascade of four reservoirs on the lower Jinsha River (Wudongde, Baihetan, Xiluodu, and Xiangjiaba) cumulatively alters downstream water temperatures. The researchers found a significant "heat source effect," where discharge temperatures have risen by up to 3.2°C and seasonal cycles have shifted by as much as 65 days, fundamentally reshaping the river's thermal regime.

What Problem Does This Paper Address?

Large dams don't just block fish; they act as massive thermal batteries. By storing water in deep reservoirs, they alter the natural temperature patterns of the river downstream—a phenomenon that can disrupt aquatic ecosystems, particularly fish spawning which relies on specific thermal cues.

While the impact of individual dams is well-documented, the cumulative effect of a "cascade" (multiple dams in a row) is much more complex. As water passes from one reservoir to the next, thermal changes can compound or counteract one another. This paper addresses the lack of long-term, quantitative evidence regarding how these four specific mega-reservoirs on the Jinsha River collectively shift water temperatures over time and distance.

Methods

The authors analyzed long-term water temperature data from five key hydrological stations (Longjie, Wudongde, Baihetan, Xiluodu, and Xiangjiaba) with records stretching back as far as 1960. To quantify the "cumulative effect," they employed three primary indicators:

  1. Baseline Shift Index ($I_{BD}$): Measures how much the average discharge temperature deviates from the natural baseline.
  2. Phase Shift Index ($I_{PS}$): Quantifies the delay in the seasonal temperature cycle (e.g., how much later the "summer peak" occurs).
  3. Extreme Amplitude Index ($I_{EC}$): Compares the range between annual maximum and minimum temperatures to natural conditions.

Additionally, they used STL (Seasonal-Trend decomposition using Loess) to separate long-term trends from seasonal fluctuations and noise, allowing them to see how the "thermal strength" of the reservoirs evolved as each new dam in the cascade was commissioned.

Datasets

The study relies on long-term measured water temperature time series from the lower Jinsha River:

  • Baseline/Input: Longjie (三) station (1999–2006).
  • Reservoir Stations: Wudongde (2020–2022), Baihetan (1989–2022), Xiluodu (2014–2022), and Xiangjiaba (1960–2022).
  • The data covers different operational phases: natural state, single-reservoir operation, and full four-reservoir cascade operation (completed in 2022).

Key Insights and Ideas

  • The "Heat Source" Effect: Contrary to some smaller reservoirs that might simply buffer extremes, this cascade acts as a net heat source. The annual average discharge temperatures at all four stations have increased compared to natural states.
  • Massive Phase Lags: The reservoirs act as a "thermal brake." By the time water reaches the final dam (Xiangjiaba), the seasonal temperature cycle is delayed by 65 days. This means the river "feels" like summer in the middle of autumn.
  • Reservoir Type Dictates Impact: Not all dams are equal.
  • Baihetan (a stable stratified reservoir) exerted the greatest influence on temperature.
  • Wudongde (mixed type) acts as a "head" reservoir that sets the initial positive thermal shift.
  • Xiangjiaba (transitional type) showed varying effects (positive or negative) depending on the operational stage of the dams upstream.
  • Slowing Rate of Change: While the cumulative impact is increasing as the reservoirs age and the cascade completes, the rate at which these changes occur is actually slowing down, suggesting the system is approaching a new, albeit artificial, thermal equilibrium.

Key Results

  • Temperature Increase: Annual discharge temperatures increased by 4.1, 2.4, 1.6, and 3.2 °C at the four respective reservoirs compared to natural conditions.
  • Reduced Warming Rate: The longitudinal warming rate of the river dropped from a natural 0.348 °C/100 km to 0.256 °C/100 km in the current cascade state.
  • Phase Delays: As of 2022, the temperature lags for Wudongde, Baihetan, Xiluodu, and Xiangjiaba were 19, 58, 49, and 65 days, respectively.
  • Amplitude Dampening: The annual temperature range (max minus min) decreased significantly, with the largest reduction being 4.1 °C at Wudongde.

Limitations and Future Work

The authors acknowledge that the full four-reservoir cascade only became fully operational in late 2022. Consequently, the data for the "complete" system is relatively short-term.

Future research directions include: * Integrating more years of joint operation data to see if the "slowing rate of change" trend holds. * Coupling these thermal findings with ecological surveys to measure the actual impact on fish biodiversity and spawning success. * Accounting for the complicating factor of global climate change, which may exacerbate the "heat source" effect of the reservoirs.

Summary updated

Abstract

金沙江下游梯级水库群在流域水温调控中扮演着关键角色,然而,对于梯级水库联合运行对下泄水温影响的认识仍然不够充分。本研究基于金沙江下游龙街(三)、乌东德、白鹤滩、溪洛渡、向家坝水文站长序列实测水温数据,从水温基线偏移、相位偏移和极值变幅3方面定量分析了金沙江下游梯级水库在不同运行阶段对沿程水温的累积影响。结果表明:(1)金沙江下游梯级水库建成后“热源效应”显著,乌东德、白鹤滩、溪洛渡、向家坝年均下泄水温逐年升高,年内水温变幅较天然水温最大减小4.1、2.4、1.6、3.2 ℃,沿程增温率由天然状态的0.348 ℃/100 km降至现阶段的0.256 ℃/100 km。(2)随着建库时间的增长,金沙江下游梯级水库水温累积程度逐渐增强,变化速度趋缓,至2022年四级水电全部投产运行后,乌东德、白鹤滩、溪洛渡、向家坝下泄水温比天然水温分别滞后19、58、49、65 d。(3)金沙江下游4个水库中白鹤滩对水温的影响最大,且不同类型水库在水温传递和累积方面呈现出独特的规律,混合型水库作为“龙头”水库时表现为正效应,稳定分层型水库表现为正效应,不稳定分层型和过渡型水库在不同阶段分别表现出正效应和负效应。本研究对于优化水资源管理、提高水电效益具有指导作用,同时为维护下游生态系统的健康提供了科学依据。
BibTeX
@article{guangyang2025water,
  author = {Hu Guangyang and Yang Zhonghua and Fengpeng Bai and Ren Yufeng and Yiming Ma and Wenhui Li},
  title = {{Water temperature's cumulative effect induced by the construction of cascade reservoirs in the lower reaches of Jinsha River}},
  journal = {Journal of Lake Sciences},
  year = {2025},
  doi = {10.18307/2025.0151},
}
Figure