| 28 Jul 2026
Limits on volcanic forcing of East Asian monsoon rainfall from the 247-year Seoul Chugugi precipitation record (1778–2024)
Abstract. The 247-year Seoul precipitation record (1778–2024), comprising 130 years of daily Chugugi rain-gauge measurements (1778–1907; Lee et al., 2024) extended by modern instrumental observations (KMA station 108, 1908–2024), constitutes the longest pre-industrial daily rainfall record in the world. Here we exploit this archive to test whether the circumglobal-teleconnection (CGT) mechanism that has been proposed to connect tropical volcanic eruptions to East Asian monsoon drought (Nie et al., 2026) imprints detectably on Seoul summer (JJA) rainfall. We apply Superposed Epoch Analysis (SEA) over 9 major eruptions (1783 Laki through 2022 Hunga Tonga) with 5,000-sample Monte Carlo bootstrap significance, continuous Morlet wavelet spectral analysis with AR(1) red-noise testing, and extended wavelet coherence between annual global VEI ≥ 4 eruption counts and Seoul JJA rainfall over 1830–2024 (N = 195). Three findings emerge. (i) The composite SEA shows no statistically significant post-eruption rainfall response at any lag from 0 to +7 yr (composite tropical-only –65 mm to +76 mm across the window; p > 0.45 at every post-eruption lag). Individual eruptions divide roughly equally between wet and dry post-eruption responses: Tambora (+298 mm), Krakatau (+132 mm), Agung (+267 mm), and Katmai (+58 mm) show net-wet years +1 to +3, while Cosiguina (–245 mm), Santa Maria (–109 mm), El Chichón (–153 mm), Pinatubo (–184 mm), and Laki (–152 mm) show net-dry. (ii) Wavelet spectral analysis reveals two significant decadal-to-multidecadal modes above the AR(1) red-noise null: a quasi-40-yr Pacific mode and a quasi-60-yr Pacific decadal mode. The interannual ENSO-like band (3–5 yr) does not exceed the null at the global-spectrum level, and the AR(1) coefficient of detrended JJA rainfall is essentially zero (−0.026), substantially weaker than previously assumed for monsoon precipitation series. (iii) Time-frequency wavelet coherence between volcanism and Seoul rainfall does not exceed the 95 % AR(1) null at any period. We do detect a modest patch of elevated coherence in the 8–16 yr band around 1900–1915 (Santa Maria–Katmai era), but no analogous patch around the recent Pinatubo–Hunga Tonga era. These findings constitute a constraint, not a confirmation: even the longest pre-industrial daily precipitation record in the world is insufficient to attribute a CGT-consistent volcanic monsoon signal at an individual mid-latitude site. We interpret this as evidence that volcanic forcing, where present in Seoul rainfall, is overwhelmed by background ENSO and Pacific-decadal variability, with event-to-event response heterogeneity likely reflecting eruption-specific stratospheric loading and background-state conditioning. The Seoul record nonetheless complements gridded paleoclimate proxies (MADA v2, RAP) by anchoring them with high temporal fidelity at a single mid-latitude site where East Asian monsoon and putative volcanic teleconnections converge.
Reviewer’s Comments
This paper utilizes a 247-year precipitation record from Seoul (1778–2024)—combining historical observations from the Gwansanggam (the Royal Observatory of the Chosŏn Dynasty, 1778–1907) with modern meteorological measurements (1908–2024)—to analyze long-term precipitation trends. By evaluating nine major volcanic eruptions with VEI≥5 during this period, the paper investigates post-eruption precipitation changes in Seoul and challenges the prevailing paradigm. The dominant consensus posits that tropical volcanic eruptions consistently trigger drought across South Asia and northern East Asia for 1–3 years post-eruption due to various climate drivers. Conversely, this paper argues that large tropical eruptions elicit no statistically significant precipitation response on the Korean Peninsula, suggesting that any volcanic forcing is offset by the El Niño–Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO).
In light of these claims, two major concerns and several minor points need serious consideration.
Major Concerns
The methodology used to select major tropical eruptions presents clear limitations. Between 1778 and 2024, the number of eruptions with VEI≥5 exceeds the nine selected in this study. Furthermore, successive small-to-moderate eruptions (VEI≤4) over a short span can collectively exert a climatic impact equal to or greater than a single large eruption.
A prominent example is the period from 1809 to April 1815, characterized by a cluster of massive and moderate eruptions:
1809: An unknown tropical eruption (VEI 6)
1812: Soufrière eruption in the Caribbean (VEI 4)
1813: Suwanose-jima eruption in the Ryukyu Islands (VEI 4)
1814: Mayon eruption in the Philippines (VEI 4)
April 1815: Tambora eruption (VEI 7)
During this six-year window, five volcanic events occurred in rapid succession, subjecting South Asia and northern East Asia to severe climatic disruptions, including abrupt cooling and extreme fluctuations in precipitation. The region did not recover from these anomalies until after 1819. In climatology and history, this period is recognized as “the last stage of the Little Ice Age” (LIA). While extensive Western literature documents these extreme climate events and human impacts, research on the Korean Peninsula includes:
Kim, S. W. (2023). Successive volcanic eruptions (1809–1815) and two severe famines of Korea (1809–1810, 1814–1815) seen through historical records. Climatic Change, 176.
Kim, S. W. (2025). Two severe famines (1809–1810, 1814–1815) in Korea during the last stage of the Little Ice Age. Climate of the Past, 21(9).
Historically, the Korean Peninsula experienced severe summer droughts during the year of an eruption, followed by recovery or torrential rainfall 2–3 years later. Although average decade-long precipitation during this LIA stage remained lower than normal, successive eruptions introduced non-linear climate variations that distorted typical post-eruption patterns. For example, following the April 1815 Tambora eruption, the peninsula did not experience a major drought; instead, it suffered record-breaking downpours and flooding the following year. Consequently, the criteria for selecting major volcanic events in this study appear somewhat inadequate.
Although the climate of the Korean Peninsula generally aligns with broader northern East Asian patterns following major tropical eruptions, significant regional variations occur along latitudinal gradients. Following the 1809 unknown eruption, precipitation dropped precipitously across the peninsula, hitting Chŏlla Province hardest, followed by Kyŏngsang Province. In contrast, regions north of Ch’ungch’ŏng Province received moderate rainfall, while P’yŏngan and Hamgyŏng Provinces showed minimal deviation. This resulted in the catastrophic "Great Famine of 1809–1810," centered in the southern provinces.
A second extreme drought struck in 1814, persisting through the summer of 1815. Known as the "Great Famine of 1814–1815," this was the most severe famine in the history of the Chosŏn Dynasty. Again, damage was concentrated in Kyŏngsang and Chŏlla Provinces, followed by Ch’ungch’ŏng Province, whereas northern provinces enjoyed bountiful harvests and even shipped surplus grain south.
These historical dynamics indicate that while volcanic forcing affects northern East Asia as a whole, local responses vary substantially by latitude. Given that the two largest famines in Chosŏn history occurred within six years and primary impacts were confined to southern Korea, the northern limit of tropical volcanic influence during the late LIA may have been around 35∘–36∘N.
Therefore, relying exclusively on precipitation data from Seoul (central Korea, 37.56∘ N) to conclude that tropical volcanic forcing has no significant climatic impact across the Korean Peninsula—or northern East Asia—presents a logical flaw.
While pre-1907 Korean instrumental data is limited to Seoul, high-quality modern meteorological records exist across the peninsula from 1908 onward. To rigorously track post-eruption precipitation behavior, the authors are recommended to disaggregate the post-1908 dataset into regional zones:
Central Region: Seoul (37.56∘ N)
Southern Region: Daegu (35.87∘ N), Gwangju (35.16∘ N), and Cheju Island (33.50∘N)
Northern Region: Pyongyang (39.0∘ N) etc.
Addressing this regional heterogeneity will significantly refine the paper's core arguments. Further elaborating on the limitations and future research directions outlined in Lines 321–324 and 336–340 would greatly enhance the scholarly contribution of this paper.
Minor Points
Abstract Length: Standard journal abstracts generally should not exceed 200 words. The current abstract contains nearly 440 words, which is more than double the standard limit. Please condense it to under 200 words while retaining only the core thesis and findings.
Eruption Intensity Differentiation [Lines 157–163]: While the study selects eruptions with VEI≥5, the magnitude variations among these nine events are substantial. For instance, the 1815 Tambora eruption produced greater forcing than all subsequent eruptions combined. Treating each eruption as an equivalent independent variable distorts the trend analysis. The authors need to account for volcanic magnitude and associated climate anomalies with finer granularity.