Large-scale Climate Modes Research Articles

Climatology of landfalling atmospheric rivers and its attribution to extreme precipitation events over Yangtze River Basin

Atmospheric rivers (ARs) have been reported to be related to Extreme Precipitation Events (EPEs). Hence, we made an assessment in the Yangtze River Basin (YRB) to understand the nexus between EPEs and ARs since the YRB has frequently encountered flooding. We threshold 6-hly fields of ERA-Interim Integrated Vapor Transport (IVT) based on the 85th percentile and a lower limit of 100 kg/m/s, and check for geometry requirements and other considerations indicative of ARs conditions. Large-scale climate modes were used to examine linkages between the frequency of ARs and the atmospheric circulation system. We recorded 75 EPEs for all 16 heavy precipitation-related flood events that occurred on days when precipitation exceeded 95% of the 3-day moving average from 1985 to 2019. Further, huge moisture transport amounts between 500 and 1000 kg/m/s were recorded some days before these identified events. Consequently, 14,347 ARs were identified, and the statistics showed that the length, width, length-width ratio, IVT, and coherent in IVT direction had medians of 3733 km, 615 km, 6.39, 350 kg/m/s, and 0.89, respectively. During most of the flood events, ARs frequencies were relatively high, with about 24% making landfall on YRB. Moreover, the deviance between the entire mean and mean composites during ARs days for the selected events in their respective study periods were in the range of 1.00–2.37 mm and 10–75 kg/m/s for precipitation and IVT, respectively, which could have resulted in the observed flood events. The positive tendencies of the NINO 3.4 and MeiV.2 indices enhanced ARs activities, thereby leading to more extreme events. The ARs distribution and associated impacts indicate a significant role of ARs in shaping the global hydrological cycle. These results could further bridge the gap to more robust hydrometeorological studies in China.

Improvement of Extreme Value Modeling for Extreme Rainfall Using Large-Scale Climate Modes and Considering Model Uncertainty

Extreme value modeling for extreme rainfall is one of the most important processes in the field of hydrology. For the improvement of extreme value modeling and its physical meaning, large-scale climate modes have been widely used as covariates of distribution parameters, as they can physically account for climate variability. This study proposes a novel procedure for extreme value modeling of rainfall based on the significant relationship between the long-term trend of the annual maximum (AM) daily rainfall and large-scale climate indices. This procedure is characterized by two main steps: (a) identifying significant seasonal climate indices (SCIs), which impact the long-term trend of AM daily rainfall using statistical approaches, such as ensemble empirical mode decomposition, and (b) selecting an appropriate generalized extreme value (GEV) distribution among the stationary GEV and nonstationary GEV (NS-GEV) using time and SCIs as covariates by comparing their model fit and uncertainty. Our findings showed significant relationships between the long-term trend of AM daily rainfall over South Korea and SCIs (i.e., the Atlantic Meridional Mode, Atlantic Multidecadal Oscillation in the fall season, and North Atlantic Oscillation in the summer season). In addition, we proposed a model selection procedure considering both the Akaike information criterion and residual bootstrap method to select an appropriate GEV distribution among a total of 59 GEV candidates. As a result, the NS-GEV with SCI covariates generally showed the best performance over South Korea. We expect that this study can contribute to estimating more reliable extreme rainfall quantiles using climate covariates.

Recent changes in temperature extremes in subtropical climate region and the role of large-scale atmospheric oscillation patterns

Understanding the recent variations in temperature extremes is crucial to anticipate the forthcoming incidences of extreme phenomena. However, knowledge of temperature extremes’ spatial and temporal patterns, as well as their links to atmospheric oscillation and topography, is scarce in Bangladesh. To this end, this research intends to analyze the spatial and temporal trends in recent extreme temperatures and their relationships with oscillation indices and the topography of Bangladesh. Daily temperature data obtained from 20 meteorological stations for 1980–2017 were employed for this purpose. Results revealed increasing trends in summer days (SU25), tropical nights (TR20), warm days (TX90p), warmest days (TXx), and warm nights (TN90p), while decline in the coldest days (TNn), cold days (TX10p), and cold nights (TN10p) was observed in Bangladesh. Spatial distribution of trends revealed an increase in SU25 and TN90p by 1.9–2.38, 2.33–2.90 days/decade, and a decrease in TX10p and TN10p by 1.7–3.3 days/decade in most regions. Besides, TR20 showed an increase of 3.22–4.17 days/decade in all sub-regions. The temperature extremes of Bangladesh showed a significant connection with multivariate ENSO index (MEI) and Sea Surface Temperature (SST). Besides, the extremes in most regions of the country showed a significant connection with Southern Oscillation Index (SOI) and Indian Ocean Dipole (IOD). The influence of atmospheric oscillation indices was more evident on cold days/nights than on warm days/nights. TN10p and SU25 also showed a significant correlation with elevation, which suggests an increase in cold night and summer day temperature with the increase in elevation in Bangladesh. Large-scale climate mode reanalysis revealed that a strong (weak) wind speed, enhancing (decreasing) geopotential height, and fast warming (cooling) over the northwestern (southeast) region have attributed to the variations in extreme temperature in Bangladesh to several extents. These findings will assist the policymakers in disaster mitigation and climate change adaptation in Bangladesh.

Evaluation of the capability of regional climate models in reproducing the temporal clustering in heavy precipitation over Europe

Global climate models (GCMs) have been used extensively as a tool to investigate future changes in heavy precipitation. However, their spatial resolution is generally too coarse for their application for decision making and for impact studies at a more local scale. To mitigate these issues, dynamical downscaling of heavy precipitation using regional climate models (RCMs) can provide information at a spatial resolution that makes their outputs locally useful. Although many studies highlighted the improvements by RCMs in reproducing the precipitation processes, much less is known about their capability in reproducing the temporal clustering of heavy precipitation. Here we use Cox regression to investigate temporal clustering in heavy precipitation driven by four dominant large-scale climate modes (Arctic Oscillation, North Atlantic Oscillation, East-Atlantic Pattern, and Scandinavia Pattern) over Europe. Results are based on high-resolution RCMs (i.e., 0.11° and 0.44°) part of the Coordinated Downscaling Experiment-European Domain (EURO-CORDEX) and use observed daily precipitation from the E-OBS as reference data. In addition to analyses at the continental scale, we consider the added value of these simulations at a regional scale. We find that RCMs are skillful at capturing the observed temporal clustering of heavy precipitation events across Europe. Moreover, the 0.11°-RCMs did not perform better than their 0.44°-version; these findings can potentially suggest that the current configuration of these models has reached an upper limit of performance in reproducing the temporal clustering, and finer resolution and modeling may be necessary to increase the realism in reproducing the processes at play.

Marine Heatwaves in the Chesapeake Bay

Prolonged events of anomalously warm sea water temperature, or marine heatwaves (MHWs), have major detrimental effects to marine ecosystems and the world's economy. While frequency, duration and intensity of MHWs have been observed to increase in the global oceans, little is known about their potential occurrence and variability in estuarine systems due to limited data in these environments. In the present study we analyzed a novel data set with over three decades of continuous in situ temperature records to investigate MHWs in the largest and most productive estuary in the US: the Chesapeake Bay. MHWs occurred on average twice per year and lasted 11 days, resulting in 22 MHW days per year in the bay. Average intensities of MHWs were 3°C, with maximum peaks varying between 6 and 8°C, and yearly cumulative intensities of 72°C × days on average. Large co-occurrence of MHW events was observed between different regions of the bay (50–65%), and also between Chesapeake Bay and the Mid-Atlantic Bight (40–50%). These large co-occurrences, with relatively short lags (2–5 days), suggest that coherent large-scale air-sea heat flux is the dominant driver of MHWs in this region. MHWs were also linked to large-scale climate modes of variability: enhancement of MHW days in the Upper Bay were associated with the positive phase of Niño 1+2, while enhancement and suppression of MHW days in both the Mid and Lower Bay were associated with positive and negative phases of North Atlantic Oscillation, respectively. Finally, as a result of long-term warming of the Chesapeake Bay, significant trends were detected for MHW frequency, MHW days and yearly cumulative intensity. If these trends persist, by the end of the century the Chesapeake Bay will reach a semi-permanent MHW state, when extreme temperatures will be present over half of the year, and thus could have devastating impacts to the bay ecosystem, exacerbating eutrophication, increasing the severity of hypoxic events, killing benthic communities, causing shifts in species composition and decline in important commercial fishery species. Improving our basic understanding of MHWs in estuarine regions is necessary for their future predictability and to guide management decisions in these valuable environments.

Relationships between large-scale climate modes and the South Atlantic Ocean wave climate

Modes of variability in ocean wave conditions are coupled to atmospheric circulation changes due to exchange of energy and momentum at the interface. Here, we explored for the South Atlantic Ocean the relations between three main climate oscillations (El Niño–Southern Oscillation [ENSO], Southern Annular Mode [SAM], and Pacific Decadal Oscillation [PDO]), four wave parameters (significant wave height [Hs], mean wave period [Tm], and zonal [Dm,x] and meridional [Dm,y] wave direction components) and wind parameters (wind speed [WS10], and zonal [u10] and meridional [v10] components). For this purpose, we regressed wind and wave parameters against the oscillation indices to create spatial composites of slope values, quantifying the correlation between wave parameters and indices. An EOF (empirical orthogonal function) analysis was also carried out to identify variability modes of wave parameters and to associate them to each climate index. The combining effects of ENSO and SAM were analysed by calculating Hs, Tm and wind speed anomalies for the periods in which the phases of these oscillations co-occur. We found important correlations not only with the dominant mode of variability, but also with secondary and even quaternary modes. For ENSO, negative correlations between the Oceanic Niño Index (ONI) and Hs, Tm, and Dm,x in the northwest part of the South Atlantic Ocean were highlighted, with a decrease (increase) of up to 8 cm of Hs per ONI unit in El Niño (La Niña) events. We established positive correlations also between ONI and these wave parameters in subtropical regions along the western African coast during austral summer, which were intensified by negative SAM. During autumn, however, we observed La Niña positive Hs anomalies for this region, which were also intensified by negative SAM. Finally, we found new, significant correlations between South Atlantic Ocean wave climate and SAM. We determined that the PDO index has negative correlations with Hs and Tm, while directional components present stronger variability.

The unidentified eruption of 1809: a climatic cold case

Abstract. The “1809 eruption” is one of the most recent unidentified volcanic eruptions with a global climate impact. Even though the eruption ranks as the third largest since 1500 with a sulfur emission strength estimated to be 2 times that of the 1991 eruption of Pinatubo, not much is known of it from historic sources. Based on a compilation of instrumental and reconstructed temperature time series, we show here that tropical temperatures show a significant drop in response to the ∼ 1809 eruption that is similar to that produced by the Mt. Tambora eruption in 1815, while the response of Northern Hemisphere (NH) boreal summer temperature is spatially heterogeneous. We test the sensitivity of the climate response simulated by the MPI Earth system model to a range of volcanic forcing estimates constructed using estimated volcanic stratospheric sulfur injections (VSSIs) and uncertainties from ice-core records. Three of the forcing reconstructions represent a tropical eruption with an approximately symmetric hemispheric aerosol spread but different forcing magnitudes, while a fourth reflects a hemispherically asymmetric scenario without volcanic forcing in the NH extratropics. Observed and reconstructed post-volcanic surface NH summer temperature anomalies lie within the range of all the scenario simulations. Therefore, assuming the model climate sensitivity is correct, the VSSI estimate is accurate within the uncertainty bounds. Comparison of observed and simulated tropical temperature anomalies suggests that the most likely VSSI for the 1809 eruption would be somewhere between 12 and 19 Tg of sulfur. Model results show that NH large-scale climate modes are sensitive to both volcanic forcing strength and its spatial structure. While spatial correlations between the N-TREND NH temperature reconstruction and the model simulations are weak in terms of the ensemble-mean model results, individual model simulations show good correlation over North America and Europe, suggesting the spatial heterogeneity of the 1810 cooling could be due to internal climate variability.

A review of the observed air temperature in the Antarctic Peninsula. Did the warming trend come back after the early 21st hiatus?

Recent changes in the near-surface air temperature (nSAT) in the Antarctic Peninsula (AP) suggests that the absence of 21st century warming on Antarctic Peninsula may be coming to end. To examine this, the long-term annual and seasonal variability of the nSAT at eight Antarctic stations located in the AP are analyzed using available data from the SCAR Reader database, complemented with data from the Chilean Weather Service (Frei and O'Higgins). An exponential filter was applied to the original annual and seasonal mean series to obtain a decadal-like variation of the nSAT. A stacked and the standardized anomaly of the nSAT record was constructed to examine the average regional behavior in the AP. Cumulative sum (CUSUM) and changepoint analysis were applied through the stacked nSAT series to highlight significant changes caused by variation in weather and climate. The CUSUM and bootstrapping analysis revealed two statistically significant breaking points during the 1978–2020 period. The first one occurred in the late nineties ending a warming period and making the beginning of a cooling period; the second one may have taken place in the mid-2010s and could mark the end of the warming pause. These trends appear to be consistent with the changes observed in the large-scale climate modes (i.e., the Antarctic Annular Mode – AAO).

MJO induced diurnal sea surface temperature variations off the northwest shelf of Australia observed from Himawari geostationary satellite

High frequency coupling between the tropical upper ocean and atmosphere at diurnal time scales is being recognised to be important for the behaviour of large scale climate modes such as the Madden Julian Oscillation (MJO). Sea surface temperature (SST) variability in the Eastern Indian Ocean warm pool and on the Northwest Shelf of Australia is strongly influenced by the MJO activity. In this study, we use the hourly Himawari-8 geostationary satellite SST data to understand the modulation of the amplitude of SST diurnal variations in the region by the MJO during two austral summer seasons of 2015–16 and 2016–17. The high temporal and spatial resolutions of the Himawari-8 SST data are able to capture the temporal and spatial variability of diurnal SST amplitude. Our results show that the MJO onset (active phase) is preceded by strong diurnal SST variations during the MJO suppressed phase. The amplitudes of the SST diurnal variations are mainly modulated by surface wind variability in the region, as denoted by their spatial and temporal correlations. The very strong El Niño event in 2015–16 may have significantly weakened the MJO activities in the region, resulting in prolonged MJO suppressed phase, high diurnal SST variation and mean SST. This study provides background knowledge for future research on how the diurnal SST variation in the region may influence the progression of the MJO events.

Coupled effects of climate teleconnections on drought, Santa Ana winds and wildfires in southern California

Projections of future climate change impacts suggest an increase of wildfire activity in Mediterranean ecosystems, such as southern California. This region is a wildfire hotspot and fire managers are under increasingly high pressures to minimize socio-economic impacts. In this context, predictions of high-risk fire seasons are essential to achieve adequate preventive planning. Regional-scale weather patterns and climatic teleconnections play a key role in modulating fire-conducive conditions across the globe, yet an analysis of the coupled effects of these systems onto the spread of large wildfires is lacking for the region. We analyzed seven decades (1953–2018) of documentary wildfire records from southern California to assess the linkages between weather patterns and large-scale climate modes using various statistical techniques, including Redundancy Analysis, Superposed Epoch Analysis and Wavelet Coherence. We found that high area burned is significantly associated with the occurrence of adverse weather patterns, such as severe droughts and Santa Ana winds. Further, we document how these fire-promoting events are mediated by climate teleconnections, particularly by the coupled effects of El Niño Southern Oscillation and Atlantic Multidecadal Oscillation.

Characterization of south central Pacific Ocean wind regimes in present and future climate for pearl farming application

In the South Pacific (SP) pearl farming atolls, wind is the main driver of lagoon water circulation, affecting dispersal and survival of pearl oyster larvae. To characterize typical wind conditions in the SP, wind regime classifications are performed from regional climate simulations using the WRF model, for present-day and for the end of the 21st century under RCP8.5 scenario conditions. At the daily time-scale, 4 regimes are identified: a trade-wind, a north-easterly, and two easterly regimes. Their characteristics are driven by large-scale circulation and climate modes of variability. In future projection, all regimes are characterized by a ~15% wind speed increase, while directions and occurrence frequencies undergo marginal changes. At the monthly time-scale that corresponds to pearl oyster pelagic larval duration, nine wind regimes are determined including three regimes with wind reversals. These regimes can be used to model typical lagoon conditions during larval dispersal.

On the role of increased CO2 concentrations in enhancing the temporal clustering of heavy precipitation events across Europe

This study examines the response of the clustering behavior in heavy precipitation events across Europe to increasing atmospheric CO2 concentrations. We focus on four large-scale climate modes [Arctic Oscillation (AO), North Atlantic Oscillation (NAO), East Atlantic (EA) pattern, and Scandinavia pattern (SCAND)] that have been found to play a significant role in controlling the occurrence of these events. We use a peak over threshold (POT) method to define daily heavy precipitation events and Cox regression as the modeling framework. We consider three experiments from the Coupled Model Intercomparison Project Phase 5 (CMIP5): (1) pre-industrial control run experiments (PI-control); (2) a 1%/year increase in CO2 from the CO2 concentration in PI control to quadrupling (1pctCO2); and (3) an instantaneous quadrupling of CO2, then holding it fixed (abrupt4 × CO2). We measure the effects of CO2 by examining whether the Cox regression coefficients in the CO2 experiments are significantly different from those in the PI control. We find that (1) the increases in CO2 are unlikely to lead to changes in the spatial patterns of relationships between climate modes and heavy precipitation; (2) the increases in CO2 are likely to lead to a strengthening of the relationship between the AO/SCAND and the occurrence of heavy precipitation events, while increases in CO2 have weaker effects on the role of NAO and EA; (3) the response to an abrupt increase in CO2 is generally stronger compared to a more gradual one; (4) the responses of the integrated vapor transport and 500-mb geopotential height to increasing CO2 provide a physical mechanism to explain the enhanced relationship between climate modes and clustering of heavy precipitation events.

Relations between Interannual Variability of Regional-Scale Indonesian Precipitation and Large-Scale Climate Modes during 1960–2007

AbstractRegional-scale precipitation responses over Indonesia to major climate modes in the tropical Indo–Pacific Oceans, namely canonical El Niño, El Niño Modoki, and the Indian Ocean dipole (IOD), and how the responses are related to large-scale moisture convergences are investigated. The precipitation responses, analyzed using a high-spatial-resolution (0.5° × 0.5°) terrestrial precipitation dataset for the period 1960–2007, exhibit differences between the dry (July–September) and wet (November–April) seasons. Canonical El Niño strongly reduces precipitation in central to eastern Indonesia from the dry season to the early wet season and northern Indonesia in the wet season. El Niño Modoki also reduces precipitation in central to eastern Indonesia during the dry season, but conversely increases precipitation in western Indonesia in the wet season. Moisture flux analysis indicates that corresponding to the dry (wet) season precipitation reduction due to the canonical El Niño and El Niño Modoki anomalous divergence occurs around the southern (northern) edge of the convergence zone when one of the two edges is located near the equator (10°S–15°N) associated with their seasonal migration. This largely explains the seasonality and regionality of precipitation responses to canonical El Niño and El Niño Modoki. IOD reduces precipitation in southwestern Indonesia in the dry season, associated with anomalous moisture flux divergence. The seasonality of precipitation response to IOD is likely to be controlled by the seasonality of local sea surface temperature anomalies in the eastern pole of the IOD.

Spring Season in Western Nepal Himalaya is not yet Warming: A 400-Year Temperature Reconstruction Based on Tree-Ring Widths of Himalayan Hemlock (Tsuga dumosa)

The Himalayan region has already witnessed profound climate changes detectable in the cryosphere and the hydrological cycle, already resulting in drastic socio-economic impacts. We developed a 619-yea-long tree-ring-width chronology from the central Nepal Himalaya, spanning the period 1399–2017 CE. However, due to low replication of the early part of the chronology, only the section after 1600 CE was used for climate reconstruction. Proxy climate relationships indicate that temperature conditions during spring (March–May) are the main forcing factor for tree growth of Tsuga dumosa at the study site. We developed a robust climate reconstruction model and reconstructed spring temperatures for the period 1600–2017 CE. Our reconstruction showed cooler conditions during 1658–1681 CE, 1705–1722 CE, 1753–1773 CE, 1796–1874 CE, 1900–1936 CE, and 1973 CE. Periods with comparably warmer conditions occurred in 1600–1625 CE, 1633–1657 CE, 1682–1704 CE, 1740–1752 CE, 1779–1795 CE, 1936–1945 CE, 1956–1972 CE, and at the beginning of the 21st century. Tropical volcanic eruptions showed only a sporadic impact on the reconstructed temperature. Also, no consistent temperature trend was evident since 1600 CE. Our temperature reconstruction showed positive teleconnections with March–May averaged gridded temperature data for far west Nepal and adjacent areas in Northwest India and on the Southwest Tibetan plateau. We found spectral periodicities of 2.75–4 and 40–65 years frequencies in our temperature reconstruction, indicating that past climate variability in central Nepal might have been influenced by large-scale climate modes, like the Atlantic Multi-decadal Oscillation, the North Atlantic Oscillation, and the El Niño-Southern Oscillation.

Dry-Season Snow Cover Losses in the Andes (18\xb0\u201340\xb0S) driven by Changes in Large-Scale Climate Modes

The Andean snowpack is the primary source of water for many communities in South America. We have used Landsat imagery over the period 1986–2018 in order to assess the changes in the snow cover extent across a north-south transect of approximately 2,500 km (18°–40°S). Despite the significant interannual variability, here we show that the dry-season snow cover extent declined across the entire study area at an average rate of about −12% per decade. We also show that this decreasing trend is mainly driven by changes in the El Niño Southern Oscillation (ENSO), especially at latitudes lower than 34°S. At higher latitudes (34°–40°S), where the El Niño signal is weaker, snow cover losses appear to be also influenced by the poleward migration of the westerly winds associated with the positive trend in the Southern Annular Mode (SAM).

Characterising spatio-temporal variability in seasonal snow cover at a regional scale from MODIS data: the Clutha Catchment, New Zealand

Abstract. A 16-year series of daily snow-covered area (SCA) for 2000–2016 is derived from MODIS imagery to produce a regional-scale snow cover climatology for New Zealand's largest catchment, the Clutha Catchment. Filling a geographic gap in observations of seasonal snow, this record provides a basis for understanding spatio-temporal variability in seasonal snow cover and, combined with climatic data, provides insight into controls on variability. Seasonal snow cover metrics including daily SCA, mean snow cover duration (SCD), annual SCD anomaly and daily snowline elevation (SLE) were derived and assessed for temporal trends. Modes of spatial variability were characterised, whilst also preserving temporal signals by applying raster principal component analysis (rPCA) to maps of annual SCD anomaly. Sensitivity of SCD to temperature and precipitation variability was assessed in a semi-distributed way for mountain ranges across the catchment. The influence of anomalous winter air flow, as characterised by HYSPLIT back-trajectories, on SCD variability was also assessed. On average, SCA peaks in late June, at around 30 % of the catchment area, with 10 % of the catchment area sustaining snow cover for > 120 d yr−1. A persistent mid-winter reduction in SCA, prior to a second peak in August, is attributed to the prevalence of winter blocking highs in the New Zealand region. In contrast to other regions globally, no significant decrease in SCD was observed, but substantial spatial and temporal variability was present. rPCA identified six distinct modes of spatial variability, characterising 77 % of the observed variability in SCD. This analysis of SCD anomalies revealed strong spatio-temporal variability beyond that associated with topographic controls, which can result in snow cover conditions being out of phase across the catchment. Furthermore, it is demonstrated that the sensitivity of SCD to temperature and precipitation variability varies significantly across the catchment. While two large-scale climate modes, the SOI and SAM, fail to explain observed variability, specific spatial modes of SCD are favoured by anomalous airflow from the NE, E and SE. These findings illustrate the complexity of atmospheric controls on SCD within the catchment and support the need to incorporate atmospheric processes that govern variability of the energy balance, as well as the re-distribution of snow by wind in order to improve the modelling of future changes in seasonal snow.

A global assessment of marine heatwaves and their drivers

Marine heatwaves (MHWs) can cause devastating impacts to marine life. Despite the serious consequences of MHWs, our understanding of their drivers is largely based on isolated case studies rather than any systematic unifying assessment. Here we provide the first global assessment under a consistent framework by combining a confidence assessment of the historical refereed literature from 1950 to February 2016, together with the analysis of MHWs determined from daily satellite sea surface temperatures from 1982–2016, to identify the important local processes, large-scale climate modes and teleconnections that are associated with MHWs regionally. Clear patterns emerge, including coherent relationships between enhanced or suppressed MHW occurrences with the dominant climate modes across most regions of the globe – an important exception being western boundary current regions where reports of MHW events are few and ocean-climate relationships are complex. These results provide a global baseline for future MHW process and prediction studies.

A classification of synoptic weather patterns linked to extreme rainfall over the Limpopo River Basin in southern Africa

In the last few decades, the Limpopo River Basin (LRB) has experienced a number of extreme rainfall events which were responsible for considerable socio-economic and environmental impacts. Most of the population here is poor and dependent on rain-fed agriculture. In order to better understand these events over the LRB, CHIRPS, 0.05° gridded rainfall data are used to identify the daily extreme events, analyse their interannual variability and examine relationships with large scale climate modes over the 1981–2016 period. Analysis of the top 20 events suggests a pattern with rainfall generally decreasing from the eastern to western parts of the basin. Typically, the highest rainfall amounts occur over the regions where there are steep topographical gradients between the mountainous regions of northeastern South Africa and the Mozambican floodplains. Almost half of the top 200 extreme events are associated with tropical extra-tropical cloud bands (48%), with tropical low-pressure systems (28%), Mesoscale Convective Systems (14%), and cut-off lows (10%) in the mid-upper atmosphere, also making sizeable contributions. The monthly distribution of the events showed that most of the events occurred during the late summer months (January–March) when tropical lows and cloud bands are more common. On interannual time-scales, most of the summers with above average number of events coincide with La Nina conditions and, to lesser extent, a positive subtropical South Indian Ocean Dipole.

Potential Predictability of Regional Precipitation and Discharge Extremes Using Synoptic-Scale Climate Information via Machine Learning: An Evaluation for the Eastern Continental United States

Abstract Current generation general circulation models (GCMs) simulate synoptic-scale climate state variables such as geopotential heights, specific humidity, and integrated vapor transport (IVT) more reliably than mesoscale precipitation. Statistical downscaling methods that condition precipitation on GCM-based, synoptic-scale climate features have shown promise in the reproduction of local precipitation. However, current approaches to climate-state-informed downscaling impose some limitations on the skill of precipitation reproduction, including hard clustering of climate modes into a discrete set of states, utilization of numerical clustering methodologies poorly suited to nonnormal data, and a tendency to focus on relationships to a limited set of large-scale climate modes. This study presents a methodology based on emerging machine learning techniques to develop global approximators of regional precipitation and discharge extremes given a suite of synoptic-scale climate state variables. Archetypal analysis is first used to define regional modes of winter and summer extreme precipitation and discharge across the eastern contiguous United States. A 2D convolution neural network (NN) is then used to predict the co-occurrence of the archetypes using 300- and 700-hPa geopotential heights, 300- and 700-hPa specific humidity, and IVT. Results suggest that 300-hPa geopotential height, 700-hPa specific humidity, and IVT yield the most reliable predictions, although with some important differences by season and region. Finally, we demonstrate that the trained activations of NN convolutional layers can be used to infer the causal pathways between synoptic-scale climate features and regional extremes.

Examining controls on peak annual streamflow and floods in the Fraser River Basin of British Columbia

Abstract. The Fraser River Basin (FRB) of British Columbia is one of the largest and most important watersheds in western North America, and home to a rich diversity of biological species and economic assets that depend implicitly upon its extensive riverine habitats. The hydrology of the FRB is dominated by snow accumulation and melt processes, leading to a prominent annual peak streamflow invariably occurring in May–July. Nevertheless, while annual peak daily streamflow (APF) during the spring freshet in the FRB is historically well correlated with basin-averaged, 1 April snow water equivalent (SWE), there are numerous occurrences of anomalously large APF in below- or near-normal SWE years, some of which have resulted in damaging floods in the region. An imperfect understanding of which other climatic factors contribute to these anomalously large APFs hinders robust projections of their magnitude and frequency. We employ the Variable Infiltration Capacity (VIC) process-based hydrological model driven by gridded observations to investigate the key controlling factors of anomalous APF events in the FRB and four of its subbasins that contribute nearly 70 % of the annual flow at Fraser-Hope. The relative influence of a set of predictors characterizing the interannual variability of rainfall, snowfall, snowpack (characterized by the annual maximum value, SWEmax), soil moisture and temperature on simulated APF at Hope (the main outlet of the FRB) and at the subbasin outlets is examined within a regression framework. The influence of large-scale climate modes of variability (the Pacific Decadal Oscillation (PDO) and the El Niño–Southern Oscillation – ENSO) on APF magnitude is also assessed, and placed in context with these more localized controls. The results indicate that next to SWEmax (univariate Spearman correlation with APF of ρ^ = 0.64; 0.70 (observations; VIC simulation)), the snowmelt rate (ρ^ = 0.43 in VIC), the ENSO and PDO indices (ρ^ = −0.40; −0.41) and (ρ^ = −0.35; −0.38), respectively, and rate of warming subsequent to the date of SWEmax (ρ^ = 0.26; 0.38), are the most influential predictors of APF magnitude in the FRB and its subbasins. The identification of these controls on annual peak flows in the region may be of use in understanding seasonal predictions or future projected streamflow changes.