This study focuses on the soil moisture characteristics and its role in supporting the continental tropical convergence zone (CTCZ) during the active phase of the monsoon. Like rainfall, land surface parameters (soil moisture and evaporation) also show intraseasonal oscillation. Furthermore, the sub-seasonal and seasonal features of soil moisture are different from each other. During the summer monsoon season, the maximum soil moisture is found over western coastal regions, central parts of India, and the northeastern Indian subcontinent. However, during active phases of the monsoon (i.e., on sub-seasonal timescales), the maximum positive soil moisture anomaly was found in northern India. Land surface characteristics (soil moisture) also play a pre-conditioning role during active phases of the monsoon over the monsoon core zone of India. When it is further divided into two boxes, the north monsoon core zone and the south monsoon core zone, it is found that the preconditioning depends on that region's soil type and climate classification. Also, we calculate the moist static energy (MSE) budget during the monsoon phases to show how soil moisture feedback affects the boundary layer MSE and rainfall. A similar analysis is applied to the model run, but it cannot show the realistic preconditioning role of soil moisture and its feedback on the rainfall as in observations. We conclude that to get proper feedback between soil moisture and precipitation during the active phase of the monsoon in the model, the pre-conditioning of soil moisture should be realistic.

How to cite: Gautam, P., Chattopadhyay, R., Martin, G., Joseph, S., and Sahai, A. K.: Intraseasonal Oscillation of Land Surface Moisture and  its role in the maintenance of land CTCZ during the active  phases of the Indian Summer Monsoon, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-572, https://doi.org/10.5194/egusphere-egu24-572, 2024.

Recent studies suggest that La Niña events can be classified into two categories: mega La Niña and equatorial La Niña. The understanding of the variations in boreal summer intraseasonal oscillation (BSISO) behaviors between such two conditions remains uncertain. Results in this work show during equatorial La Niña summers, in conjunction with the more adequate intraseasonal column-integrated moisture anomalies, the weaker intraseasonal outgoing longwave radiation anomalies are observed over the western North Pacific (WNP) at 3 pentads lag of the peak phase for the Maritime Continent (MC) BSISO events than during mega conditions. Such changes are closely linked with the different propagation features, specifically northwestward and northeastward propagations under mega and equatorial conditions respectively. The distinct propagations under these two conditions could be partly explained by the background column-integrated moisture anomalies. Under equatorial conditions, the less sufficient background moisture anomalies over the tropical western Pacific (WP), in comparison to mega conditions, suppress the activities of the BSISO and its northwestward propagation here. Meanwhile, the enhanced moisture anomalies over the northwestern MC and its surrounding area (NWMC) facilitate the northeastward propagation. Under mega conditions, the background moisture anomalies over the tropical WP are not significant. The southward moisture anomaly gradient over the NWMC hinders the meridional northward propagation and makes some BSISO activities move to the tropical WP region, performing the zonal westward propagation as a whole. The moisture budget and multi-scale interaction diagnoses also emphasize the significant role of the propagation change in the moisture tendency difference averaged over the WNP. Moreover, the extratropical circulation anomalies associated with the MC BSISO events are also discussed. These findings provide new insights into BSISO activity and offer potential improvements for subseasonal forecast.

How to cite: Cao, C. and Wu, Z.: Distinct changes in boreal summer intraseasonal oscillation over the western North Pacific under mega and equatorial La Niña conditions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1244, https://doi.org/10.5194/egusphere-egu24-1244, 2024.

Subseasonal prediction of extremes has emerged as a top forecast priority but remains a great challenge. In this work, we explored two physical modes controlling the subseasonal variation and prediction of land cold extremes over Eurasia: the so-called North Atlantic Oscillation (NAO) and the Eurasian Meridional Dipole mode (EMD). The ECMWF model has shown its skill in predicting the Eurasian land cold extremes 2-4 weeks in advance mainly because of the skillful prediction of NAO and EMD. Further, we separated these observed events into the good prediction and poor prediction groups for those two modes to reveal the potential factors influencing the subseasonal prediction of land cold extremes. It is found that the good prediction group has a stronger initial amplitude and longer persistence, while the poor prediction group has a relatively weaker initial amplitude but rapid intensification. For EMD, the predictability is mainly due to the skillful prediction of the Ural blocking which is further traced back to the stratospheric variations.  

How to cite: Xiang, B.: The window of opportunity for subseasonal land cold extreme prediction over Eurasia  , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1376, https://doi.org/10.5194/egusphere-egu24-1376, 2024.

The hypothesis that predictability depends on the atmospheric state in the planetary-scale low-frequency variability in boreal winter was examined.We first computed six typical weather patterns from 500-hPa geopotential height anomalies in the Northern Hemisphere using self-organizing map (SOM) and k-clustering analysis. Next, using 11 models from the subseasonal-to-seasonal (S2S) operational and reforecast archive, we computed each model’s climatology as a function of lead time to evaluate model bias. Although the forecast bias depends on the model, it is consistently the largest when the forecast begins from the atmospheric state with a blocking-like pattern in the eastern North Pacific. Moreover, the ensemble-forecast spread based on S2S multimodel forecast data was compared with empirically estimated Fokker– Planck equation (FPE) parameters based on reanalysis data. The multimodel mean ensemble-forecast spread was correlated with the diffusion tensor norm; they are large for the cases when the atmospheric state started from a cluster with a blocking-like pattern. As the multimodel mean is expected to substantially reduce model biases and may approximate the predictability inherent in nature, we can summarize that the atmospheric state corresponding to the cluster was less predictable than others.

How to cite: Inatsu, M., Matsueda, M., Nakano, N., and Kawazoe, S.: Prediction Skill and Practical Predictability Depending on the Initial Atmospheric States in S2S Forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1548, https://doi.org/10.5194/egusphere-egu24-1548, 2024.

This study investigates the influence of the boreal summer intraseasonal oscillation (BSISO) on 10-30-day summer rainfall anomalies in Southwestern China (SWC) under the effects of Qinghai-Tibetan Plateau monsoon (QTPM) based on ERA5 reanalysis data and CN05.1 precipitation in 1981-2018. The results show that the 10-30-day rainfall anomalies in SWC have significant and joint feedback to variation of the second component of BSISO (BSISO2) and QTPM at lagging strong (weak) BSISO events by 0-12 days. Their lagged causal linkage and corresponding physical processes have been revealed by causal effect networks and composite analyses, which are most significant at 4-day and 12-day lag. Simultaneously, BSISO2 can induce wetter 10-30-day rainfall over southern SWC by motivating water vapor transport from the Bay of Bengal towards Yunnan province. More importantly, BSISO2 can modulate a northwest-propagating wave train from the western north Pacific towards SWC at the upper troposphere by vertical wave energy transport, which blocks the wave train propagating from the Lake Balkhash to east China–Japan most significantly at a 4-day lag and leads to drier eastern SWC. The process can be influenced by QTPM significantly which leads to the response of 10-30-day rainfall over SWC with lags of 0-12 days. Specifically, same-phase QTPM can trigger more active wave train propagation from high-latitude while opposite-phase QTPM enhances the low-latitude wave energy transport. The interference then facilitates baroclinic structure over eastern SWC at lagging 12 days with positive precipitation anomalies for same-phase events and negative precipitation for opposite-phase events.

How to cite: Yang, L., Chen, H., and Wang, S.: The joint effects of the boreal summer intraseasonal oscillation and Qinghai-Tibetan Plateau monsoon on the precipitation over Southwestern China , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2784, https://doi.org/10.5194/egusphere-egu24-2784, 2024.

The Madden-Julian Oscillation (MJO) is the dominant intraseasonal variability of eastward propagating atmospheric disturbances in the tropics. From its vast impacts on the sub-seasonal extreme events and predictability, the mean states controlling the MJO activity have been investigated. For example, the robust relationship between the Quasi-Biennial Oscillation (QBO) and the MJO has been suggested in the past several years. In the easterly QBO winters, the MJO exhibits stronger activity than the westerly QBO winters. 
Our study suggests another crucial factor that affects the MJO: a meridional humidity gradient of the atmospheric column in the vicinity of the Maritime Continent. With the change in the shape of the column humidity distribution, MJO variance shows a robust interannual modulation regardless of the QBO. The northward (southward) extension of the moisture increases (decreases) the mean state meridional humidity gradient, which leads to MJO development (decay) over the MC with increasing (decreasing) horizontal moisture advection. This robust relationship between mean state humidity and MJO activity is investigated in the CMIP6 models as two aspects: i) interannual variation of MJO and ii) future change in MJO. Both simulated MJO activities are largely affected by the mean state MHG, supporting the robust role of mean state moisture on the MJO shown in the observations. The results of this study provide a further understanding of seasonal MJO activity and sub-seasonal predictability.MJO activity and sub-seasonal predictability.

How to cite: Kang, D., Kim, D., and Kang, S.-Y.: Robust Relationship between Mean State Moisture and Interannual MJO Activity in Observations and CMIP6 Models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2862, https://doi.org/10.5194/egusphere-egu24-2862, 2024.

The precipitation over the eastern Tibetan Plateau (ETP, here defined as 29°–38°N, 91°–103°E) usually exhibits significant subseasonal variation during boreal summer. As the hot spot of land-air interaction, the influences of ETP surface soil temperature (T_soil) on the local precipitation through subseasonal land-air interaction are still unclear but urgently needed for improving subseasonal prediction. Based on station and reanalysis datasets of 1979–2018, this study identifies the evident quasi-biweekly (QBW) (9–30 days) periodic signal of ETP surface T_soilvariation during the early summer (May–June), which results from the anomalies of southeastward propagating mid-latitude QBW waves in the mid-to-upper troposphere. The observational results further show that the maximum positive anomaly of precipitation over the ETP lags the warmest surface T_soil by one phase at the QBW timescale, indicating that the warming surface T_soil could enhance the subseasonal precipitation. The numerical experiments using the WRF model further demonstrate the effect of warming surface T_soil  on enhancing the local cyclonic and precipitation anomaly through increasing upward sensible heat flux, the ascending motion, and water vapor convergence at the QBW timescale. In contrast, the effect of soil moisture over the ETP is much weaker than T_soil  at the subseasonal timescale. This study confirms the importance of surface T_soil over the ETP in regulating the precipitation intensity, which suggests better simulating the land thermal feedback is crucial for improving the subseasonal prediction.

How to cite: Qi, X., Yang, J., Xue, Y., Bao, Q., Wu, G., and Ji, D.: Subseasonal Warming of Surface Soil Enhances Precipitation Over the Eastern Tibetan Plateau in Early Summer, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3148, https://doi.org/10.5194/egusphere-egu24-3148, 2024.

Global warming is accelerating drought onset, causing more frequent flash drought events. These events occur at the subseasonal timescale in which rapid decreases in root-zone soil moisture (RZSM) increase risks of crop failure, wildfire, and heat stress globally. However, forecasting soil moisture and flash droughts at lead times beyond 2 weeks remains a significant challenge. Recently, machine learning methods with historical reanalysis data have shown improved forecast accuracy compared to state-of-the-art numerical weather prediction methods, but they can only produce skillful forecast within 10 days. Here we show that a convergence forecast model combining a deep learning approach with subseasonal retrospective forecasts (reforecast) from numerical models produces skillful subseasonal soil moisture and flash drought forecasts at lead times beyond 2 weeks. We train a deep learning architecture on combinations of reanalysis and reforecast from 2000 to 2015 and validate results during the testing period from 2018 to 2019. The subseasonal forecast skill of soil moisture of the convergence forecast model is much higher than those of current state-of-the-art numerical forecast models, deep learning bias corrected numerical forecast models, or the reanalysis-based deep learning models, which showed no skill after 2 weeks lead time. The convergence model also showed significantly improved performance for predicting flash droughts compared to the original or deep learning bias corrected numerical forecast models or reanalysis-based deep learning models.  A permutation analysis indicates that reanalysis precursors and soil moisture reforecast at lead times within 2 weeks both contribute significantly to the forecast skill at longer lead times. The convergence forecast model provides accurate and efficient subseasonal soil moisture and flash drought forecasting and is promising for accurately forecasting key variables and extreme events at the subseasonal timescale.

How to cite: Lesinger, K. and Tian, D.: Converging Deep Learning and Numerical Prediction for Skillful Subseasonal Soil Moisture and Flash Drought Forecasting, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3194, https://doi.org/10.5194/egusphere-egu24-3194, 2024.

Despite current global warming due to increasing greenhouse gases, severe cold winters have devastated the East Asia in recent decades. Efforts are being made to predict cold events using dynamic models and physically-based statistical models. In this study, we explore the potential predictability of the East Asian winter surface temperature by establishing a multiple linear regression model based on three precursors of time-evolved preconditions: 1) autumn Arctic sea-ice loss, 2) northern Eurasian sea level pressure pattern, and 3) the El Niño-Southern Oscillation (ENSO). Reduced autumn Arctic sea-ice was favorable for extreme cold events in the East Asia. Furthermore, the autumn Arctic sea-ice loss was accompanied by cyclonic circulations over northern Eurasia in November, which could have led to cold anomalies over the East Asia in the late winter. The preconditioning deep convection in La Niña events is a well-known indicator of exerted atmospheric wave propagation, resulting in cold winters over the East Asia. We suggested here that by combining Arctic sea-ice, atmospheric circulations, and ENSO, the predictability of East Asian winter surface temperature variability could be improved.

How to cite: Jang, Y.-S., Lim, H.-G., Jun, S.-Y., and Kug, J.-S.: Arctic sea ice loss and La Niña as precursors of extreme East Asian cold winters, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3737, https://doi.org/10.5194/egusphere-egu24-3737, 2024.

The predictability of certain extreme weather events can exceed the traditional two weeks by considering the boundary conditions. Targeted observations in sensitive areas on Arctic sea ice concentration (SIC) can improve the extended-range (4 pentads) forecast skills of long-lasting and strong Ural blocking (UB). The sensitive areas are determined based on the SIC optimally growing boundary errors, obtained by the conditional nonlinear optimal perturbation method. The sensitive areas are mainly located in the Barents Sea, Greenland Sea, and Okhotsk Sea. The results of observing system simulation experiments for 8 UB cases indicate that the targeted observations can remarkably improve the prediction skills of UB in the 4th pentad. Targeted observations have a positive effect on 75% of 160 experiment members, reduce 35% forecast errors of the 4th pentad mean blocking index, and perform even better when the original forecast errors are greater. Further diagnosis shows that targeted observations contribute to more accurate SIC boundary conditions in the Barents Sea, Greenland Sea, and Okhotsk Sea and reduce temperature errors in the lower and middle troposphere. It further results in well-described westerly winds in the Ural region and its adjacent regions, corresponding to the more skillful predictions of blocking circulations. The above results supply a theoretical base for the design of Arctic SIC observations and more skillful extended-range predictions for mid-latitude extreme weather.

How to cite: Gao, Y., Mu, M., and Dai, G.: Targeted Observations on Arctic Sea Ice Concentration for Improving Extended-range Prediction of Ural Blocking, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3796, https://doi.org/10.5194/egusphere-egu24-3796, 2024.

The summer northern annular mode (NAM) variability plays a crucial role in the summer climate variability and extremes of the Northern Hemisphere. In this study, we report a significant negative correlation between the March NAM and summer NAM during 1979–2022 and reveal the role of the spring stratosphere in this seasonal linkage. Particularly, it is found that the negative phase of March NAM features a strong meridional shear in the extended-North-Atlantic jet, which tends to generate planetary scale Rossby waves that propagate upward and poleward into the stratosphere. This increased stratospheric planetary wave activity in March transitions to weakened wave activity in May, leading to positive zonal wind anomalies in the polar stratosphere in May, extending downward to the troposphere in June and promoting the formation and persistence of positive summer NAM. The results provide both statistical and dynamical evidence for the role of the spring stratosphere in connecting the spring and summer circulation. 

How to cite: Xu, X., Wang, L., Wang, T., and Chen, G.: The role of stratospheric processes in the trans-seasonal connection between spring and summer northern annular modes, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3798, https://doi.org/10.5194/egusphere-egu24-3798, 2024.

This study uses the ECMWF 46-day ensemble to evaluate the subseasonal forecasts of tropical cyclones (TCs) in the western North Pacific, including TC formations, tracks, intensity, and precipitation forecasts. TC formations and the subsequent tracks are objectively detected in both real-time forecasts and also the 20-year ECMWF reforecasts. Additionally, a spatial-temporal track clustering technique is utilized to group similar vortex tracks in the 101-member real-time forecasts for operational application. The forecast verification focuses on evaluating the influence of large-scale environmental factors on TC forecast skills during weeks 1-4, such as the Western North Pacific Summer Monsoon (WNPSM), Madden Julian Oscillation (MJO), and Boreal Summer Intraseasonal Oscillation (BSISO). The Precision-Recall (PR) curve is used to represent the imbalanced TC data instead of the Receiver Operating Characteristic (ROC) curve. Better TC forecast skills are observed if model initialized on MJO Phases 6 and 7 for the week-1 forecasts, and on MJO Phases 4 and 5 for the weeks 2 and 3 forecasts. Also, TC forecast skills are better if the cumulative percentage of the WNPSM index (Wang et al. 2001) is larger than 60%. This study also investigats the TC precipitation forecast skill around Taiwan area.

The evaluation results obtained from this study has been integrated into the TC Tracker 2.0 system developed by Central Weather Administration (CWA). The system can generate a "Subseasonal TC Threat Potential Forecast" product to assist in disaster mitigation and water resources management for the Water Resources Agency. More details about the subseasonal TC forecast verifications and applications will be presented in the meeting

How to cite: Tsai, H.-C., Hsu, H.-Y., Lo, T.-T., and Chen, M.-S.: Verifications of Week-1 to Week-4 Tropical Cyclone Forecasts in the Western North Pacific from the ECMWF 46-Day Ensemble, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4272, https://doi.org/10.5194/egusphere-egu24-4272, 2024.

Considering the significant differences in the rainfall characteristics over East Asia between the early [May–June (MJ)] and late [July–August (JA)] summer, this study investigates the subseasonal predictability of the rainfall over East Asia in early and late summer, respectively. Distinctions are obvious for both the spatial distribution of the prediction skill and the most predictable patterns, that is, the leading pattern of the average predictable time (APT1) between the MJ and JA rainfall. Further analysis found that the distinct APT1s of MJ and JA rainfall are attributable to their different predictability sources. The predictability of the MJ rainfall APT1 is mainly from the boreal intraseasonal oscillation signal, whereas that of the JA rainfall APT1 is provided by the Pacific–Japan teleconnection pattern. This study sheds light on the temporal variation of predictability sources of summer precipitation over East Asia, offering a possibility to improve the summer precipitation prediction skill over East Asia through separate predictions for early and late summer, respectively.

How to cite: Li, X.: Subseasonal Predictability of Early and Late Summer Rainfall Over East Asia, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4955, https://doi.org/10.5194/egusphere-egu24-4955, 2024.

Summer monsoon precipitation over the Bay of Bengal (BoB) has pronounced intraseasonal variability (ISV), which has a close relationship to the local intraseasonal sea surface temperature (SST). Before heavy precipitation, intraseasonal SST in the BoB often has a warm anomaly and propagates northward, which drives the atmosphere and tends to trigger the convection. Besides the local air-sea interaction, the ISV of SST in the Arabian Sea (AS) also has an effect on the precipitation over the BoB. Results show that a prominent heavy precipitation usually occurs when the warm intraseasonal SST anomaly appears early in the AS and moves northward prior to that emerges in the BoB. The warm SST anomaly in the AS affects the sea level pressure and then trigger a southwestly wind anomaly in the center of AS. This wind anomaly promotes the wind convergence moving northward from the southern tip of Indian peninsula to the north India and northern BoB, which directly influence the vertical moisture advection and finally the precipitation. Understanding this process will be helpful to improve the predictive skill of the ISVs during the Indian Summer Monsoon.

How to cite: Xi, J.: Influence of Intraseasonal Variability of Sea Surface Temperature in the Arabian Sea on the Summer Monsoon Precipitation Over the Bay of Bengal, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5422, https://doi.org/10.5194/egusphere-egu24-5422, 2024.

The Navy Earth System Prediction Capability (ESPC) is the Navy’s coupled ocean-atmosphere-sea ice model.  The current version of the Navy ESPC has 16 ensemble members and been operational since August 2020. The Navy ESPC has known biases in Madden-Julian Oscillation (MJO), which has a too strong amplitude and too fast propagation speed. During boreal winter, the MJO in the Navy ESPC is too strong due to biases in the vertical motion, which supports larger vertical moisture advection.  The MJO is too strong in this season due to excessive evaporation in the western Pacific supporting moistening to the east of the MJO convective center.  In this study, we examine the boreal winter MJO in the operational Navy ESPC ensemble.  We use process oriented diagnostics to explore the local and remote sources of biases that drive good and poor MJO forecasts. 

MJO forecasts are split into those that are well predicted and those that are poorly predicted.  Individual MJO events are tracked following Chikira (2014), using Hovmöllers of MJO filtered OLR averaged between 10N and 10S.  The MJO forecast performance is determined by comparing the forecasted MJO to the observed MJO based on the magnitude of the maximum amplitude of the MJO, the phase speed, duration of the event, and the location of the MJO convection.  Using the moisture mode framework, we examine the maintenance and propagation of moisture anomalies to identify how the local and remote sources of error affect MJO skill.  We use a moisture budget analysis to diagnose and understand the difference between the forecasts that performed well and those that performed poorly.  Additionally, we examine the effects that these forecast errors in the MJO have on extratropical cyclones, surface winds, and clouds in the Navy ESPC and how biases in the extratropics affect the skill of MJO-teleconnections.

How to cite: Rushley, S., Janiga, M., and Reynolds, C.: Local and remote sources of error inMJO forecasts in the Navy ESPC , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6452, https://doi.org/10.5194/egusphere-egu24-6452, 2024.

Recently, it has been noticed that weather and climate changes over the Arctic and mid-latitude regions may have influenced the particulate matter concentrations and haze over East Asia. Among the various weather and climate conditions and climate indices could be an important factor in affecting variation of particulate matter (PM) concentrations. In this study, we examined the long-term changes in the sea ice cover, soil moisture, near-surface temperature and its link with the lower atmospheric circulation over Arctic and mid-latitude from 1950 to 2022, using modern reanalysis datasets. Long-term analyses show negative trends in sea ice cover over the Arctic and positive trends in near-surface temperature and SST, implying atmospheric stagnant and variation of PM concentration. Additionally, climate indices, related to teleconnection between the Arctic region and mid-latitude, co-related with understanding air quality. Based on climate indices, we have developed the air quality prediction model for reflecting variations in weather and climate conditions. Therefore, the findings in this study can likely be used for actual prediction systems based on long-term weather measurement datasets over the Arctic region.

Acknowledgment: This research was supported by a National Research Foundation of Korea Grant from the Korean Government (MSIT ; the Ministry of Science and ICT) (NRF- 2023M1A5A1090715).

How to cite: Park, J.-M., Lee, D., Kim, K., Kim, S., Lee, G., and Lee, K. H.: Weather and Climate conditions over the Arctic and mid-latitude regions affecting air quality, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7386, https://doi.org/10.5194/egusphere-egu24-7386, 2024.

Seasonal predictions are essential in mitigating damage to people and nature as a result of climate change and extreme events by improving timely decision-making particularly for water and irrigation management. The newly constructed Grand Ethiopian Renaissance Dam, located in the Blue Nile (BN) Basin in Ethiopia at the border to Sudan, increases the urgency of optimized transboundary water management and improved seasonal predictions. However, the global seasonal forecasting systems have known limitations such as biases and drifts. Specifically at regional level, such as in the highlands of Ethiopia, the seasonal predictions need accurate post-processing. Recent developments have shown the large potential of Deep Learning (DL) applications to improve weather and climate predictions. The goal of this study is to improve the global seasonal forecasting system SEAS5 of ECMWF specifically for the BN Basin using DL approaches such as conventional Convolutional Neural Networks (CNN) or more advanced Adaptive Fourier Neural Operators (AFNO). We present first results for improving and downscaling SEAS5 global seasonal precipitation forecasts in the BN Basin with a particular emphasis on ensemble generation and calibration. The neural networks are trained with ERA5-Land-reanalysis data as a ground-truth, which has a higher resolution than SEAS5 (~9km compared to ~36km). This additional downscaling step allows us to consider the high variations in precipitation intensities in the Ethiopian highlands. The results show that the applied DL models have high potential in improving forecasting scores such as the continuous ranked probability skill score. They therefore allow for improved timely decision-making for water management in the transboundary BN Basin.

How to cite: Wiegels, R., Glawion, L., Polz, J., Chwala, C., Weber, J. N., Schober, T. C., Lorenz, C., and Kunstmann, H.: Deep Learning improved seasonal forecasts for the Blue Nile Basin, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11457, https://doi.org/10.5194/egusphere-egu24-11457, 2024.

We explored the objective methods to improve long-range forecasting through enhanced forecast skills and integrated forecast information. The objective process we used in this study includes the selection of monitoring factors for more reliable monthly seasonal forecasts. Therefore, we chose the three most significant monitoring factors, i.e., ENSO, snow cover over Eurasia Continent and Arctic sea ice. We first examined the effect and response of the monitoring factors on the boreal winter temperature in South Korea. To improve the information related to the ENSO in seasonal forecasting, the impact of the tropical precipitation which act as an oceanic ENSO forcing was investigated. As one of the important monitoring factors for boreal winter temperature prediction, we analyzed the availability of the index describing austral Eurasian snow cover. We also analyzed the usage of Arctic conditions for predicting monthly temperature for boreal winter. We then investigated how well the effect and response of the factors are simulated in the operational seasonal models. Finally, the link between observation-based monitoring factors and model-based prediction is proposed for objective forecasting.

How to cite: Kim, O., Im, S.-H., and Kim, G.: Improved long-range forecasts in South Korea through integrated forecast information, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14585, https://doi.org/10.5194/egusphere-egu24-14585, 2024.

Given the limited skill of precipitation forecasts, the question arises to what extent ensemble forecasting systems can be used for early warning systems that require longer lead times, such as drought early warning. In this study, we use ECMWF’s IFS extended range forecasts, statistically downscaled to a 2 km grid encompassing Switzerland, to quantify the spatially and seasonally stratified predictability of several precipitation statistics. Consistent with existing analyses we find the predictability of extratropical instantaneous precipitation to be limited to week 1. However, when considering accumulated precipitation and the standardized precipitation index (SPI) forecasts, which is commonly used for drought management, the forecasts are skillful well into week 3. This extension in predictability horizon is attributed to the characteristic of accumulated precipitation, which is less sensitive to differences in timing of precipitating systems. The enhanced predictability of SPI enhances the utility of extended range forecasts for monthly drought forecasts. We discuss the practical applicability of these findings in the context of the new Swiss drought early warning and monitoring platform, planned for operations in 2025. Leveraging the enhanced predictability of SPI, this platform stands to benefit from our research outcomes, providing stakeholders with tools for proactive drought management and response strategies. 

How to cite: Imamovic, A., Büeler, D., Pyrina, M., Humphrey, V., Spirig, C., Moret, L., and Domeisen, D.: How far in advance can we skillfully predict meteorological drought indices?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15134, https://doi.org/10.5194/egusphere-egu24-15134, 2024.

In recent years, deep learning (DL) models have been shown to be able to make competitive forecasts of El Niño Southern Oscillation (ENSO). In most cases, due to the short observational record, the outputs of global circulation models (GCM) are used to train DL models. However, GCMs themselves show biases when modeling ENSO dynamics, such as the lack of phase-locking behavior, shifted precipitation trends, or missing El Niño - La Niña asymmetry. The biases of the GCMs are likely inherited by the DL models during pre-training, raising the question of how we can obtain unbiased DL ENSO models while pre-training on GCM output. In this study, we contend that a possible solution to correct these biases is to use well-established domain adaptation methods, which allow DL models to account for shifts in data distribution between training and validation data sets. In particular, we use a ConvLSTM network trained on CESM2 simulations where we first use a supervised objective to fine-tune our model to reanalysis data. Secondly, we employ test-time training to adapt our model for the domain shift between CESM2 and reanalysis data. This study serves as a first step toward comparing domain adaptation techniques for data-driven seasonal-to-annual DL models in a limited data regime.

How to cite: Rodriguez, M.: Domain adaptation for deep learning ENSO forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15210, https://doi.org/10.5194/egusphere-egu24-15210, 2024.

Droughts, prolonged heat-waves, heavy precipitation events and large-scale flooding - the last years have demonstrated that global climate change is already hitting hard in many places of the Earth. This, inevitably, leads to increased water stress that requires a more sustainable and timely water management across scales. In particular, for optimized use of water resources for irrigation or hydropower generation, it is essential to know their expected availability in the coming months all over the world. This sub-seasonal to seasonal temporal domain, from weeks to months ahead, is addressed by seasonal forecasting systems such as SEAS5, developed by the European Centre for Medium-Range Weather Forecasts (ECMWF). These systems have the potential to provide essential data for enhancing water management practices. Without a bias correction though, the data exhibit a notable deficiency in skill. We have shown for several regions of the world that the “Bias Correction and Spatial Disaggregation” method (BCSD) can improve the forecasting skill substantially. Our next step is now to expand our efforts from the regional to the global scale, i.e., to provide the BCSD-forecasts for the entire globe. Here, the challenge lies in significantly reducing the computational demand for the bias correction: Presently, the BCSD requires several days to execute on a global scale. However, if such forecasts should be used as decision support, a timely provision is crucial.

We therefore present a method to achieve this task: The utilization of fixed Cumulative Distribution Functions (CDFs) rather than their recalculation for each pixel has the potential to enhance the computational efficiency of the bias correction. This approach not only significantly reduces the required data volume but also improves accessibility. To further achieve transferability of the system, we also demonstrate the performance of this system in a containerized environment. Our goal is to achieve a globally corrected SEAS5 forecasts within a time frame of ideally less than one day. With the provision of these bias-corrected data in near-real time, better estimations become available for direct utilization by water managers or as input for subsequent modeling processes.

How to cite: Weber, J. N., Lorenz, C., Schober, T., Wiegels, R., and Kunstmann, H.: Global bias-corrected seasonal forecasts: Towards efficient and near real-time solutions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16022, https://doi.org/10.5194/egusphere-egu24-16022, 2024.

The seasonal cycle (SC) anomalies of winter surface air temperature (SAT) over China mainly include three modes: consistent changes throughout the winter, inverse changes in the early and late winter, and opposite changes in the southern and northern China, respectively.  The positive EOF1 phase (i.e., uniformly warming throughout winter) can be attributed to global warming, especially in the North Atlantic and tropical Pacific. The EOF2 is mainly related to the dipole sea surface temperature (SST) pattern in the North Atlantic. In the early winter, the Rossby wave originating from North Atlantic strengthens Ural blocking high (UBH) and Siberian high (SH) in the early winter, resulting in cold SAT anomalies in most of China. While the large-scale zonal circulation with weakened SH has transformed SAT over China into a warm state in the later winter. The EOF3 can be attributed to the tripole SST in the North Atlantic and El Niño-like SST pattern in the tropical Pacific. In December, the Rossby wave train originating from the mid-latitudes of the North Atlantic Ocean enhances cold air activity in the Northern Hemisphere, causing cold SAT anomalies in Northeast China, while the dominating southerly winds in southern China cause warm SAT anomalies. In the late winter, the large-scale circulation resembles negative AO phase, resulting in the northerly winds and cold SAT anomalies in the northern China. Meanwhile, the anomalous anticyclonic circulation in the Northwest Pacific causes warm SAT anomalies in southern China. Therefore, the combined effects of tropical and extratropical SST should be considered when predicting interannual variability of winter SAT anomalies over China.

How to cite: Yu, M.: Diversity of seasonal cycle anomalies of surface air temperature in winter over China , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16753, https://doi.org/10.5194/egusphere-egu24-16753, 2024.

Combining ensemble forecasts from different models into a multi-model ensemble (MME) have been shown to improve forecast skill at different time-scales, including the sub-seasonal to seasonal (S2S) one. Here, we investigate a new method to build such MME based on barycenter.

Recognizing ensemble forecasts as discrete probability distributions, we work directly in the probability distribution space. This allows us to use existing tools in this space, and in particular the concept of barycenter. The barycenter of a collection of distributions (or the ensemble forecasts here) is the distribution that best represents them, based on a given metric. The barycenter can thus be seen as the combination of these distributions, and so used to build a MME. We compare two barycenters based on different metrics: the L2 and the Wasserstein distances. The Wasserstein distance corresponds to the cost of the optimal transport between two distributions and has interesting properties in the distribution space. We compare it to the L2-barycenter which is in fact shown to be equivalent to the well-known “pooling” MME method (i.e. the concatenation of the different ensembles members). Another interesting point of the barycenters is that they allow you to give different weights to the models and so to easily build a weighted-MME. The weights have an important impact on the skill of the MMEs. We are thus optimizing the weights by learning them from the data using cross-validation on the forecasts.

The two barycenter-based MMEs are applied to the combination of the models from the S2S project’s database. Their performances are evaluated for the prediction of weekly 2m-temperature during European winter with respect to different metrics. As a proof of concept, we first start with the combination of two models, namely the European Centre Medium-Range Weather Forecasts (ECMWF) and the National Center for Environmental Prediction (NCEP) models. We show that the two MMEs are generally able to perform as well or better than both the single-models, but that the best combination method depends on the chosen metric. We then extend the barycenter approach to the combination of more models, of which we will discuss preliminary results.

How to cite: Le Coz, C., Tantet, A., Flamary, R., and Plougonven, R.: A barycenter-based approach for the multi-model ensembling of subseasonal forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17318, https://doi.org/10.5194/egusphere-egu24-17318, 2024.

Predictive skills of coupled sea-ice/ocean and atmosphere models are limited by the chaotic nature of the atmosphere. Assimilation of observational information on ocean hydrography and sea ice allows to obtain a coupled-system state that provides a basis for subseasonal-to-seasonal ocean and sea-ice forecast (Mu et al., 2022). However, if the atmosphere is not additionally constrained, the quasi-random atmospheric states within an ensemble forecast lead to a fast divergence of the ocean and sea-ice states, degrading the system’s performance with respect to the sea ice forecasts. As reported previously, imposing an additional constraint by nudging large-scale winds to the ERA5 reanalysis data (Sánchez-Benítez et al., 2021; Athanase et al., 2022) improves predictive skills of the AWI Coupled Prediction System (AWI-CPS, Mu et al. 2022) with regard to sea ice drift (Losa et al., 2023). Here we provide results based on a much more extensive set of ensemble-based data assimilation experiments spanning the time period from 2002 to 2023 and a series of long forecast experiments over 2010 – 2023, initialized in four different seasons. We compare the performance of forecasts initialized from two sets of data assimilation experiments, with and without atmospheric wind nudging. The additional relaxation of the large-scale atmospheric circulation to the ERA5 reanalysis data for the initialization leads to reasonable atmospheric forecast skill on weather timescales: Despite the simple technique, the coarse resolution compared to NWP systems, and the limited optimization efforts, 10-day forecasts of the 500 hPa geopotential height are about as skillful as the best performing NWP forecasts were about 10 –15 years ago. Among other aspects, this leads to significantly improved subseasonal-to-seasonal sea-ice concentration and thickness forecasts.

Athanase, M., Schwager, M., Streffing, J., Andrés-Martínez, M., Loza, S., and Goessling, H.: Impact of the atmospheric circulation on the Arctic snow cover and ice thickness variability , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5836, https://doi.org/10.5194/egusphere-egu22-5836, 2022.

Losa, S. N., Mu, L., Athanase, M., Streffing, J., Andrés-Martínez, M., Nerger, L., Semmler, T., Sidorenko, D., and Goessling, H. F.: Combining sea-ice and ocean data assimilation with nudging atmospheric circulation in the AWI Coupled Prediction System, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-14227, https://doi.org/10.5194/egusphere-egu23-14227, 2023.

Mu, L. , Nerger, L. , Streffing, J. , Tang, Q. , Niraula, B. , Zampieri, L., Loza, S. N. and Goessling, H. F. (2022): Sea‐Ice Forecasts With an Upgraded AWI Coupled Prediction System , Journal of Advances in Modeling Earth Systems, 14 (12) . doi: 10.1029/2022ms003176

Sánchez-Benítez, A. , Goessling, H. , Pithan, F. , Semmler, T. and Jung, T. (2022): The July 2019 European Heat Wave in a Warmer Climate: Storyline Scenarios with a Coupled Model Using Spectral Nudging , Journal of Climate, 35 (8), pp. 2373-2390 . doi: 10.1175/JCLI-D-21-0573.1

How to cite: Loza, S., Athanase, M., Mu, L., Streffing, J., Sánchez-Benítez, A., Andrés-Martínez, M., Nerger, L., Semmler, T., Sidorenko, D., and Goessling, H.: Improving daily-to-seasonal sea ice forecasts of the AWI coupled prediction system with sea-ice and ocean data assimilation and atmospheric large-scale wind nudging., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18260, https://doi.org/10.5194/egusphere-egu24-18260, 2024.

Accurate subseasonal forecast of East Asian summer monsoon precipitation (EASM) is pivotal, impacting the livelihoods of billions. However, the proficiency of state-of-the-art subseasonal-to-seasonal (S2S) models in forecasting precipitation remains constrained. We developed a convolutional neural network regression model, harnessing the more reliably predicted atmospheric variables from dynamic models to enhance their forecast skills for precipitation. The outcomes of the CNN model are promising: a 12% increase in accuracy and a 10% reduction in RMSE for precipitation forecast at the lead time of one week. The predictive skill of dynamic models for atmospheric variables shows a significant correlation with the performance of the CNN model. Ablation experiments on various predictors reveal that xx is the most influential factor affecting the CNN model's performance.

How to cite: Jiahui, Z. and Liu, F.: Improving sub-seasonal forecasting of East Asian monsoon precipitation with deep learning, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19908, https://doi.org/10.5194/egusphere-egu24-19908, 2024.

The year-to-year varying annual evolutions of the stratospheric polar vortex (SPV) have an important downward impact on the weather and climate from winter to summer and thus potential implications for seasonal forecasts. This study constructs a parametric elliptic orbit model for capturing the annual evolutions of mass-weighted zonal momentum at 60° N (MU) and total air mass above the isentropic surface of 400 K (M) over the latitude band of 60–90° N from 1 July 1979 to 30 June 2022. The elliptic orbit model naturally connects two time series of a nonlinear oscillator. As a result, the observed coupling relationship between MU and M associated with SPV as well as its interannual variations can be well reconstructed by a limited number of parameters of the elliptic orbit model. The findings of this study may pave a new way for short-time climate forecasts of the annual evolutions of SPV, including its temporal evolutions over winter seasons as well as the spring and fall seasons, and timings of the sudden stratospheric warming events by constructing its elliptic orbit in advance.

How to cite: Secor, M., Yu, Y., Sun, J., Cai, M., and Luo, X.: A Parametric Model of Elliptic Orbits for Annual Evolutions of Northern Hemisphere Stratospheric Polar Vortex and Their Interannual Variability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20725, https://doi.org/10.5194/egusphere-egu24-20725, 2024.

Extratropical cyclones influence midlatitude surface weather directly via precipitation and wind and indirectly via upscale feedbacks on the large-scale flow. Biases in cyclone frequency and characteristics in medium-range to sub-seasonal numerical weather prediction might therefore hinder exploiting the potential predictability on these timescales. We thus, for the first time, identify and track extratropical cyclones in 21 years (2000 – 2020) of sub-seasonal ensemble reforecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF) in the Northern Hemisphere in all seasons. Overall, the reforecasts reproduce the climatology of cyclone frequency and life-cycle characteristics qualitatively well up to six weeks ahead. However, there are significant regional biases in cyclone frequency, which can result from a complex combination of biases in cyclone genesis (locally and upstream), size, location, lifetime, and propagation speed. Their magnitude is largest in summer, with the strongest deficit of cyclones of up to 15% in the North Atlantic, relatively large in spring, and smallest in winter and autumn. Moreover, the reforecast cyclones are too deep in both ocean basins during most seasons, although intensification rates are captured well. An overestimation of cyclone lifetime and differences between the native spatial resolutions of the reforecasts and the verification dataset might explain this intensity bias in some cases, but there are likely further so far unidentified processes involved. While the patterns of cyclone frequency and life cycle biases often appear in lead time weeks 1 and 2, their magnitudes typically grow further at sub-seasonal lead times and, in some cases, saturate in weeks 5 and 6 only. Most of the dynamical sources of these biases thus likely appear in the early medium range, but biases on longer timescales probably contribute to their further increase with lead time. Our study provides a useful basis to identify, better understand, and ultimately reduce biases in the large-scale flow and in surface weather in sub-seasonal weather forecasts. Given the considerable biases during summer, when sub-seasonal predictions of precipitation and surface temperature will become increasingly important, this season deserves particular attention for future research.

How to cite: Sprenger, M., Büeler, D., and Wernli, H.: Northern Hemisphere extratropical cyclone biases in ECMWF sub-seasonal forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14701, https://doi.org/10.5194/egusphere-egu24-14701, 2024.