International Journal of Climatology最新文献

The Atlantic Niño, a dominant climate phenomenon in the tropical Atlantic with substantial impacts on surrounding continents, is significantly modulated by the Indian Ocean dipole (IOD). However, the relationship between these two climate modes exhibits pronounced interdecadal variations over the past decades, and the underlying mechanisms remain elusive. With a focus on the possible causes of this non-stationarity, we find that neither the IOD's own forcing strength nor the concurrent influence of Pacific ENSO alone can fully explain these interdecadal variations. Observational analyses and numerical experiments with CESM1.2.2 demonstrate that Atlantic preconditions play a crucial role in shaping the IOD–Atlantic Niño relationship. During the period when the IOD–Atlantic Niño relationship is weak, the positive phase of the boreal autumn IOD often coincided with cold sea surface temperature (SST) anomalies in the equatorial Atlantic in the preceding summer. These cold Atlantic conditions suppress the development of an Atlantic Niño in subsequent months. In contrast, during the period when the IOD–Atlantic Niño relationship is significant, the boreal autumn IOD generally occurred alongside normal Atlantic preconditions, favouring an in-phase relationship between the boreal autumn IOD and the following winter Atlantic Niño. Furthermore, we find that the differences in the Atlantic preconditions are related to the type of ENSOs. Continuing ENSOs, developing from the previous winter and prevalent during the weak IOD–Atlantic Niño relationship period, induce pronounced tropical Atlantic preconditions in the following summer. In contrast, emerging ENSOs, developing later in the year and more common during the strong relationship period, are associated with normal Atlantic preconditions. These findings enhance our understanding of the IOD–Atlantic Niño teleconnection and have significant implications for the seasonal prediction of the latter.

Drought prediction plays a vital role in mitigating the adverse effects of climate variability and ensuring sustainable water resource management. In order to classify drought in the northern Iranian provinces of Golestan, Mazandaran and Guilan, this study assessed the effectiveness of machine learning models such as random forest (RF), AdaBoost, decision tree (DT) and transformer. Historical climate data were preprocessed to include lagged features and statistical aggregates for tree-based models, while normalised data were directly used for the transformer model to capture temporal dependencies. F1-score, recall, accuracy and precision were used to evaluate the model's performance. The results revealed that RF consistently outperformed other models across all regions, demonstrating superior accuracy, precision and recall. AdaBoost followed closely, while DT provided moderate performance. The transformer model showed limited effectiveness, particularly in Guilan and Mazandaran. Optimal hyperparameters were determined for each model, ensuring robust evaluation and providing a benchmark for future studies. The findings underscore the effectiveness of RF in drought prediction and highlight the regional variability in model performance. This research emphasises the importance of model selection and tuning in achieving reliable predictions and offers insights into the application of machine learning techniques for drought monitoring. Future research should explore advanced models, such as deep learning or hybrid approaches, and consider additional climatic variables to enhance predictive accuracy further.

Drought significantly affects agriculture and ecology in the Horn of Africa (HOA), whereby livelihoods largely depend on rainfed farming. This study aims to analyze drought propagation and its impacts on vegetation and crop productivity, with a specific focus on recovery dynamics. Over the period 2000–2022, we developed and integrated a suite of physically based remote sensing indices, including the Drought Propagation Index (DPI), Crop Stress Index (CSI), Soil Moisture Deficit (SMD), Water Deficit Index (WDI), and Drought Recovery and Rate Index (DRRI), into a novel framework. The performance of this integrated framework was further evaluated against the conventional Standardised Precipitation Index (SPI) to validate its ability for capturing drought propagation and agricultural impacts. The findings identify the eastern and southeastern HOA as major drought hotspots, experiencing severe droughts 70%–100% of the time and exhibiting worsening temporal trends. SPI was strongly correlated with DPI (r = 0.67, p < 0.05), thus proving to be reliable. Vegetation indices showed significant reductions during droughts, while DPI positively correlated with NDVI at 0.56 and EVI at 0.54. Crop production was reduced by 30%–35% in Somalia, especially for maize and sorghum, whereas Ethiopia showed more resistance because of irrigation. The mean recovery time exceeded 2 months during the 2010 and 2016 droughts in southeastern HOA, indicating low resilience, whereas northern areas recovered faster. This framework offers practical recommendations for drought mitigation, drought-resistant crops, and adaptive resource management to deal with vulnerabilities.

The impact of the amplitude characteristic of the intra-seasonal oscillation (ISO) on water vapour transport is examined for the March–May (MAM) and September–November (SON) seasons. The ISO temporal indices are constructed from daily anomalies of the first two principal components of the empirical orthogonal function (EOF) applied on outgoing longwave radiation anomalies, filtered for 25–70 days. A threshold method subsequently applied to the normalised ISO amplitude using the first two EOF components is used to identify extreme events associated with ISO peaks. This allows identifying 129 (119) strong ISO events (sISOs) and 39 (40) weak ISO events (wISOs) during the MAM (SON) season. The findings revealed in MAM that, during sISO events, the southern part of the domain experiences a strengthening of the moisture convergence on the Atlantic coast, to the south-east, associated with a westerly moisture transport. In addition, sISO induces negative anomalies in the horizontal component of moist static energy (MSE) advection. wISO are characterised by a predominance of significant moisture divergence to the east of the domain and in the centre of the Congo Basin, accompanied by moisture transport from the Indian Ocean. This event generates positive moisture advection anomalies, fuelled by a zonal influx of MSE. In contrast, the two examined ISO events exhibit opposing patterns of moisture transport during the SON season. Specifically, a dipole structure emerges with divergence anomalies in the north coinciding with convergence anomalies in the south during the sISO. Analysis of the zonal components of moisture transport indicates a reinforcement during wISO events in both seasons, due to the reinforcement of the African Easterly Jet (AEJ). During wISOs, stronger low-level westerlies facilitate increased moisture import from the Atlantic Ocean. Furthermore, compared to sISO events, wISOs are associated with more intense northern and southern components of the AEJ. Finally, uncentred pattern correlation coefficients between strong and weak ISO events and the different phases of the Madden-Julian Oscillation (MJO) vary depending on the season and the specific MJO phase.

Climate change has significantly amplified the occurrence of extreme weather events globally, yet characterising and attributing record-breaking temperature extremes – an especially severe category within such phenomena – is difficult. This study analyses the statistical feature of record-breaking temperature extremes based on daily surface air temperature data from 2299 meteorological stations across China spanning from 1960 to 2023. The results indicate that summer record-breaking high-temperature events occur more frequently than theoretically predicted, whilst winter record-breaking low-temperature events exhibit the opposite pattern. Notably, post-2020 summers demonstrate a more pronounced acceleration in high-temperature record-breaking frequency compared to preceding periods. Additionally, a detrended variance-adjusted model is used to quantify how global climatic non-stationarity modulates the evolution of local daily record-breaking temperature extremes. The results show that climate-driven trends account for 10%–30% of daily record-breaking temperature events, with variance changes contributing merely 2%. These findings quantitatively elucidate the contribution of global climate change to the observed intensification in summer record-breaking high-temperature events and concomitant attenuation of winter record-low temperature extremes in recent decades, establishing a theoretical reference for the attribution of changes in record-breaking events under non-stationary climate conditions.

The Loess Plateau Region (LPR) in China, identified as one of the world's most severely affected areas by soil erosion, contains fragile ecosystems highly susceptible to climate variability. This study examines the long-term trends and abrupt shifts in temperature and precipitation extremes across the LPR from 1982 to 2023, offering a comprehensive assessment of climate change in this sensitive region. Homogenised daily data for the mean (T _mean), maximum (T _max) and minimum (T _min) temperatures, along with precipitation (PPT), were collected from 160 meteorological stations. Data quality control and homogenisation were performed using RHtestsV3. The datasets were interpolated onto a 0.75° × 0.75° grid via ordinary Kriging in ArcGIS. Monotonic trends were analysed using the Mann–Kendall test and Sen's slope, with iterative pre-whitening to mitigate autocorrelation bias. Abrupt shifts were detected using the Mann–Whitney U test, with significance assessed via Z-values and Monte Carlo simulations. We found a significant regional warming of annual T _mean at 0.0177°C/year (≈0.71°C since 1982). Annual T _max changed little (0.002°C/year, non-significant), whereas T _min decreased overall (−0.051°C/year), despite spring–summer T _min increases and winter T _min cooling (−0.018°C/year). Consequently, the diurnal temperature range widened, especially in winter. The strongest warming occurred in winter and spring, with over 83% of grid cells exhibiting significant trends (p < 0.05). Precipitation increased significantly (4.81 mm/year; 200 mm over the study period), predominantly during the rainy season (+3.02 mm/year). Spatial variability was notable, where 41% of grid cells showed precipitation increases, while 13% exhibited declines, particularly in the southeastern LPR. Abrupt shifts affected 27% of temperature and 18.5% of precipitation grid cells, predominantly around 1993–1997 and 2009–2012; however, these shifts were localised and not statistically significant at the regional scale. Persistent warming and moderately increasing precipitation characterise the LPR's recent climate, underscoring the necessity for continued monitoring and adaptive strategies to address the impacts of temperature and precipitation extremes in this vulnerable region.