No longer on schedule, the pattern Is breaking apart: The Loss of Seasonal Synchrony in a Sub-Arctic River System Under Warming Climate

Jalali Shahrood, Abolfazl; Ahrari, Amirhossein; Torabi Haghighi, Ali

doi:10.5194/egusphere-2025-2982

Preprints

https://doi.org/10.5194/egusphere-2025-2982

Preprints

07 Nov 2025

| 07 Nov 2025

No longer on schedule, the pattern Is breaking apart: The Loss of Seasonal Synchrony in a Sub-Arctic River System Under Warming Climate

Abolfazl Jalali Shahrood, Amirhossein Ahrari, and Ali Torabi Haghighi

Abstract. Climate warming is altering the timing and synchrony of snow and ice processes across northern river systems, yet long-term shifts in their seasonal dynamics remain insufficiently resolved. Here, we analyze a 57-year daily record (1966–2023) from the River Oulankajoki in northeastern Finland to characterize freeze-up, break-up, snow accumulation and melt, and key atmospheric temperature transition points. Using a process-based detection tool, we identify significant advances in spring-related events, including snow melt, ice break-up, and the seasonal shift from cold to warm temperatures. In contrast, autumn transitions such as freeze-up and snow onset exhibit higher year-to-year variability and no consistent trends. The durations of cold season, ice cover, and snow melt periods have shortened, while warm and open-water seasons have lengthened. Moreover, the temporal gap between atmospheric warming and surface responses has increased in spring but contracted in autumn. These findings suggest not only a shift in seasonal timing but also a growing desynchronization between atmospheric conditions and cryo-hydrological processes, with implications for Arctic river ecology, ice forecasting, and flood risk under continued climate change.

Received: 23 Jun 2025 – Discussion started: 07 Nov 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Abolfazl Jalali Shahrood, Amirhossein Ahrari, and Ali Torabi Haghighi

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-2982', Anonymous Referee #1, 05 Dec 2025
This manuscript applies the RiTiCE (River Ice Timing Characteristics and Extremes) framework to 57 years (1966–2023) of daily discharge, air temperature, and snow depth data from the River Oulankajoki, a boreal/sub-Arctic river in northeastern Finland. Building on earlier RiTiCE work that focused mainly on break-up, the authors extend the framework to detect six phase change timing (PCT) features (FUD, BUD, SBD, SMD, and two temperature transition points TTP⁻ and TTP⁺) and derive associated seasonal durations (cold and warm seasons, ice cover, open water, snow cover, no-snow period). The authors then analyze long-term trends, correlations among PCT features and seasonal periods, and temporal lags between atmospheric transitions and cryo-hydrological responses. A central qualitative conclusion is the asymmetry between autumn and spring. The order and timing of autumn features (SBD, TTP⁻, FUD) exhibit substantial year-to-year variability, whereas spring features (TTP⁺, BUD, SMD) retain a highly stable sequence. This behavior is interpreted as a loss of seasonal synchrony driven by climate warming.
Although this manuscript falls within the scope of TC and is based on a valuable data set, it requires substantial revision before it can be considered for publication. My main concerns are:
The central narrative relies on the sequence and relative timing of the six PCT markers. A key and potentially confusing result is that in some years FUD occurs earlier than both SBD and TTP⁻. This is counter-intuitive if one understands “freeze-up” as the date when the river is first observed to be completely ice-covered.

The manuscript states in Sect. 2.3 that FUD and BUD are hydrological signatures, distinct from IAHR visual definitions. However, later sections routinely interpret FUD as freeze-up in the physical sense. The paper needs to be more precise and transparent here, especially given the unusual orderings of events.

I suggest adding an explicit clarification such as “FUD is a discharge-based proxy for the onset of the winter low-flow regime, which is typically associated with sustained ice cover in this river, but does not necessarily coincide with the date when the channel is first completely ice-covered in a visual sense.”

In Fig. 7, the authors should explicitly discuss the years when FUD precedes SBD and/or TTP⁻. It would help to show one or two example years (similar to Fig. 6) with T, Q, S and all six PCT markers overlaid, and to guide the reader through how an early FUD can arise physically, for example a very dry autumn, early stabilization of baseflow, delayed continuous snow cover at the nearby terrestrial station, and a late zero-crossing in the cumulative temperature curve.

I think the authors need to acknowledge more explicitly the limitations of the hydrological proxy and to demonstrate that these cases are physically plausible and reasonably rare, for instance X out of 57 years, rather than artefacts of the detection algorithm.

The phrase “loss of seasonal synchrony” is prominent in the title, abstract, and conclusions, but there is no clear quantitative definition of synchrony and no consolidated figure that directly supports this claim. What exactly does "seasonal synchrony" mean here in this study?

The study relies on one air and snow station (Kuusamo Kiutaköngäs, about 500 m from the river gauge) and one discharge gauge to characterise the whole sub-Arctic river system. This is understandable given data availability, but it has direct consequences for the interpretation of the PCT features. SBD and SMD are based on terrestrial snow depth at the weather station rather than snow on the river. Vegetation and wind redistribution can lead to systematic differences between station snow and river ice or river-adjacent snow conditions.

Air temperature is also a single-point measurement, whereas freeze-up, break-up, and hydrological responses integrate the entire upstream basin. I suggest that in Sect. 4, when discuss cases where FUD precedes SBD or TTP⁻, or where snow and ice duration diverge, 他和authors should explicitly mention this spatial representativeness issue as one plausible explanation.

Specific comments:
Line 11: Calling RiTiCE a “process-based detection tool” may be misleading. It is a statistical rule-based method using thresholds on Q, T, and S, not a physical energy-balance or hydrodynamic model.

Line 95: At present the manuscript moves into the Methods in a descriptive way and does not clearly state two or three guiding scientific questions. I suggest adding a short paragraph with explicit research questions, for example “Specifically, we ask how the timings of six key PCT features have changed over 1966–2023, how the duration of cold, warm, ice-covered and snow-covered seasons have been reorganised, and how the lags between atmospheric transitions and surface hydrological and cryospheric responses have evolved.”

Line 104: The water year is defined from 1 October to 30 September. Based on Fig. 4, negative temperatures can already occur in October. For some colder water years this choice might prevent capturing the full evolution of Ct within a water year. It would be helpful to comment on this point. In addition, the figures linked in Appendix A are currently provided in emf or eps format. If possible, providing tif versions would make them easier to view.

Equation 1: the authors define DVD as ΔQ_t = Q_t+1 −Q_t. Mathematically, using Q_t −Q_t−1 would only shift indices by one day, so the stable segment and FUD/BUD would be essentially unchanged. However, from a physical perspective it may be more intuitive to define “change from day t−1 to t” instead of “from t to t+1”.

Equations 5 and 6: authors treat any S_t > 0 as “snow present” in the CSP algorithm. Please state the measurement resolution. This is especially relevant for short early/late snow events and could affect SBD/SMD in marginal years.

Section 2.5: It is not entirely clear why authors do not directly use the more classical and widely used FDD and TDD metrics. A short explanation of why the ZCAT approach is preferred here would be helpful.

Line 225: “The manuscript states that it provides “one of the most comprehensive long-term assessments”. However, authors have already published related work in Water titled “RiTiCE: River Flow Timing Characteristics and Extremes in the Arctic Region”, where the DVD method was developed and applied to 100-year discharge series in sub-Arctic rivers. I suggest softening the current claim to something like “a detailed 57-year assessment for a well-monitored sub-Arctic river”.

Line 228: Please explain why 1966 was chosen as the reference year. For example, is it simply the first year of the record, or was it selected because it is particularly representative?

Figure 9: The figure is not very clear in its current layout and would benefit from being rearranged or redrawn. In addition, important information is missing, such as the Sen slope and the Mann–Kendall p value for each variable.
Citation: https://doi.org/10.5194/egusphere-2025-2982-RC1
RC2:
'Comment on egusphere-2025-2982', Anonymous Referee #2, 30 Mar 2026
The authors are commended for their work and analysis presented which made for an interesting read. The manuscript provides an analysis of a 57-year daily record (1966-2023) of the River Oulankajoki using a process-based detection tool building on the River Ice Timing Characteristics and Extremes (RiTiCE) tool previously developed by the Authors. Analysis presented is based on identified Phase Change Timing (PCT) features of freeze-up dates (FUD), break-up dates (BUD), snow-build up days (SBD), snow melting day (SMD), and two temperature transition point (TTP) metrics. The results highlight asymmetric seasonal changes to parameters, with those defining spring-based processes showing statistically significant trends relative to the selected baseline year of 1966. Several correlations were discussed amongst the PCT metrics. The results are of significance to the scientific community for understanding long-term trends in river ice processes in sub-arctic/arctic systems and fit within the scope of The Cryosphere. There are several concerns however regarding the presented results which warrant significant changes to the manuscript prior to publication.

General comments:
The definitions of FUD and BUD based solely on discharge are problematic. While they may apply fairly well to this river system, the authors suggest a scalability of the RiTiCE framework, which, would may be problematic for rivers with frequent mid-winter freeze-thaw events, increases in discharge, or regulated rivers (e.g. hydropeaking). Further, under ice discharge values are notoriously challenging to accurately measure. It is unclear what post-processing methods (if any) have been used on the winter-time discharge data presented since 1966. Do the authors have visual ice phenology data for the river that can be used to compare to the discharge-based delineations?

The authors do not need to reintroduce acronyms in each section. They should be introduced once and used thereafter. Figures are often confusing to understand. Figures should always be introduced in text before they appear- this was missing for several figures.

Section 3.2 is confusing. It is unclear what is meant by ranks and order and how the presented figure and analysis follow the presented claim: “A central finding of this study is the asymmetry between spring and autumn transitions. The sequence of seasonal cryo-hydrological features revealed a clear separation between early- and late-year events (Fig. 7).”

Results need to be better contextualized. One gauge station is used to represent a moderately sized river system with decently large basin. This should be considered when asserting conclusions and attempting to generalize to larger Arctic and sub-arctic river systems.

Several large claims are made throughout the paper which, in my opinion, are not justified based on the results presented. Examples include L324, L334-335, L350-351, and L369-370. On the latter: the results presented in the study do not support a “fundamentally reorganized” cryo-hydrological system" as the authors' argue.

The title has too much jargon and is inappropriate for the presented analysis. It is recommended to just use “Loss of Seasonal Synchrony in a Sub-Arctic River System Under Warming Climate” or something similar.

Specific comments:
L95: The scalability of “RiTiCE” is questionable for regions that experience mid-winter melting and increased discharge using the given definitions of the metrics for BUDs and FUDs, SMD and SBD, and TTP^+/-.
L122-124: “Instead, BUD refers explicitly to the timing of the river ice break-up day which is defined as the day when the river’s flow begins to rise after a long period of low and stable discharge, corresponding to the start of ice break-up and the transition toward spring flow conditions”
This would only apply to a particular subset of unregulated rivers in which mid-winter increases in discharge (or hydropeaking in the regulated case) are not expected. Your statements on L212-215 regarding shifting seasonality in discharge values point to potential problems in defining your metrics in the approach outlined.
L130- DVD used as daily variation in discharge but is defined as Daily Value Difference elsewhere. Additionally, is the discharge used in this study the mean daily or max daily? What about post-processing corrections?
DVD definition: perhaps instead of the standard deviation, a more robust metric such as an IQR or 75^th percentile could be used to handle distribution skewness.
Section 2.5- what is the advantage here of using a “Zero-Referenced Cumulative Area Transition (ZCAT)” rather than a traditional Cumulative freezing degree days approach but correcting for temperatures greater than 0°C? It is confusing why a new metric needed to be defined here.
L227- remove brackets for PCT
L228-229: I would suggest rephrasing the sentence “The 1966 reference…” It has not been shown in this study that 1966 conditions were under stable cryo-hydrological conditions. The 1966 year is simply used as a baseline and should be treated as such without generalization.
L231- “These results…” Provide references for this statement.
L240: No specifics are presented for the full period. The reader simply points to “full-year characterizations for all 57 years are provided in Appendix A.” Include a brief description of the full period in the text. For example, what was the average FUD, BUD, SMD, etc. To strengthen section 3.1, it is perhaps more prudent to use a reference period (e.g. 5-10 years) rather than one particular year which may have conditions outside the climate normal.
L295- It is interesting how your BUD, SMD, and TTPs showed statistically significant trends but your air temperatures did not (discussed on L293-296). How do you explain this? Perhaps looking at monthly air temperature trends for the should periods of freeze-up and break-up would show significant trends, as has been shown elsewhere in the literature.
Figure 9 is rather difficult to read. Text is too small and trends are difficult to interpret.
L312-313: You state that spring PCT metric trends are from consistent warming, but your air temperature trends do not agree with this statement. Please explain this claim (see previous comment also).
L350-351: The results presented from one gauged station (presented here) cannot be extrapolated to speculate on most sub-Arctic systems.
Citation: https://doi.org/10.5194/egusphere-2025-2982-RC2

Abolfazl Jalali Shahrood, Amirhossein Ahrari, and Ali Torabi Haghighi

Viewed

Total article views: 410 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
218	163	29	410	26	29

HTML: 218
PDF: 163
XML: 29
Total: 410
BibTeX: 26
EndNote: 29

Views and downloads (calculated since 07 Nov 2025)

Month	HTML	PDF	XML	Total
Nov 2025	85	48	14	147
Dec 2025	42	37	7	86
Jan 2026	24	30	2	56
Feb 2026	25	16	1	42
Mar 2026	36	30	3	69
Apr 2026	6	2	2	10

Cumulative views and downloads (calculated since 07 Nov 2025)

Month	HTML	PDF	XML	Total
Nov 2025	85	48	14	147
Dec 2025	42	37	7	86
Jan 2026	24	30	2	56
Feb 2026	25	16	1	42
Mar 2026	36	30	3	69
Apr 2026	6	2	2	10

Viewed (geographical distribution)

Total article views: 400 (including HTML, PDF, and XML) Thereof 400 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 11 Apr 2026

Short summary

We analyzed 57 years of river, snow, and temperature records from a sub-Arctic watershed in Finland to understand how seasonal freezing and thawing have changed due to climate warming. Our findings show that spring events like snowmelt and ice break-up are happening earlier, while autumn changes remain unpredictable. This loss of seasonal coordination affects river behavior, water resources, and future planning under ongoing climate change.


Total:	0
HTML:	0
PDF:	0
XML:	0