Seasonal Prediction Model Research Articles

Leveraging crop yield forecasts using satellite information for early warning in Senegal

Agricultural losses driven by climate variability and anthropogenic pressures have severely impacted food security in Senegal. There is a crucial need to generate early warning signals for the upcoming season to enhance food security in response to the sudden climate shocks like drought. In this study, we investigated the spatial distribution of maize and groundnut using factor analysis with a principal component approach. We aimed to identify suitable predictors of crop yields for the development of a seasonal yield prediction model. Subsequently, multi-regression analysis was performed to predict crop yield based on various combinations of satellite-derived vegetation and climate (rainfall) datasets as well as agronomic data from Senegal's 40 districts between 2010 and 2021. Studies revealed a strong correlation between seasonal rainfall (May to September) and crop yield: a 10–20% decline in rainfall can lead to crop losses. The accuracy of the yield prediction model, built on the best performing scenarios for each district based on monsoon onset, duration, and planting time, exceeded 0.5 (R-squared) for all districts when combining rainfall and normalized difference vegetation index (NDVI) data. The model prediction accuracy varied between 0.6 and 0.8 for major crop growing areas. The study emphasizes that refining the yield prediction model using machine learning techniques can improve its accuracy and enable its implementation in early warning systems. This enhanced capability could bolster Senegal's resilience to climate change by aiding decision-makers and planners in developing more effective strategies to ensure food security.

Large model biases in the Pacific centre of the Northern Annular Mode due to exaggerated variability of the Aleutian Low

AbstractThe Northern Annular Mode (NAM) is traditionally defined as the leading empirical orthogonal function (EOF) of mean sea‐level pressure (MSLP) anomalies during winter. Previous studies have shown that the Pacific centre‐of‐action of the NAM is typically more amplified in models than in reanalysis. Here, we analyse the NAM in hindcasts from nine seasonal prediction models over 1993/1994–2016/2017. In all the models, the Pacific centre‐of‐action is much larger than in reanalysis over that period, during which the NAM and the North Atlantic Oscillation (NAO) are almost indistinguishable. As a result, the NAM in the models is correlated with Aleutian Low variability around four times more strongly than in reanalysis. We show that this discrepancy can be explained primarily by the amplitude of Aleutian Low variability, which is on average 17% higher in models than in reanalysis, with a secondary effect from a stronger correlation between the Aleutian Low and NAO. When the NAM is computed using zonally averaged MSLP, the Aleutian Low amplitude does not influence the pattern directly. Instead, the amplitude of the Pacific centre‐of‐action is governed primarily by the correlation between the Aleutian Low and NAO, reducing the apparent Pacific biases in models. While the two methods yield almost identical results in reanalysis, the large Aleutian Low biases result in differences when applied to model data. Modifying the MSLP statistically to alter the Aleutian Low amplitude reveals that the spatial pattern of the traditionally defined NAM is highly sensitive to Aleutian Low variability, even without modifying the correlation between the Aleutian Low and NAO. Hence, the NAM in models may not be as biased as the traditional method would suggest. We therefore conclude that the traditional EOF method is unsuitable for defining the NAM in the presence of highly amplified Aleutian Low variability, and encourage the use of the zonal‐mean method.

Performance of weather research forecasting model for seasonal prediction of precipitation over Indonesian maritime continent

The precipitation over the Indonesian Maritime Continent (IMC) is one of the most challenging atmospheric parameters to predict accurately. The Weather Research Forecasting (WRF) model used in this study produces overestimated predictions of precipitation intensity compared to satellite data. Therefore, it is necessary to modify the model output to minimize the bias between predictions and satellite observations. The modification includes adjusting the dynamic model parameters and settings to better reflect atmospheric conditions in the IMC region, applying bias correction techniques through a linear scaling method, aligning the monthly average of the model's output with observational data, and conducting statistical analysis in several areas within the IMC. This procedure has significantly reduced the biases and is considered acceptable for each satellite area. Based on the results of the statistical analysis and by applying the precipitation threshold criteria, the accuracy of the predictions in each observation area is quite good, ranging from 0.59 to 1.0. Precipitation with a threshold of 50 mm/day or higher exhibits good accuracy, with a minimum value of 0.59 for RI and a maximum value of 0.97 for RIII and RV. On the other hand, precipitation with a threshold of 20 mm/day or higher demonstrates very good accuracy, with values of 0.97 for RI and 1.00 for RIII to RVI.

Suitability of the CICE sea ice model for seasonal prediction and positive impact of CryoSat-2 ice thickness initialization

Abstract. The Los Alamos Community Ice CodE (CICE) sea ice model is being tested in standalone mode to identify biases that limit its suitability for seasonal prediction, where it is driven by atmospheric forcings from the NCEP Climate Forecast System Reanalysis (CFSR) and a built-in mixed-layer ocean model in CICE. The initial conditions for the sea ice and mixed-layer ocean are also from CFSR in the control experiments. The simulated sea ice extent agrees well with observations during the warm season at all lead times up to 12 months, in both the Arctic and the Antarctic. This suggests that CICE is able to provide useful sea ice edge information for seasonal prediction. However, the model's initial conditions have ice that is too thick in the Beaufort Sea, resulting in excessive ice extent in the Arctic at 6-month lead forecasts and errors in ice volume at all lead times when compared to available observations. To address this limitation, additional CS2_IC experiments were conducted, where the Arctic ice thickness was initialized using CryoSat-2 satellite observations while keeping all other initial fields the same as in the control experiments. This reduced the positive bias in the ice thickness in the initial conditions, leading to improvements in both the simulated ice edge and the ice thickness at the seasonal timescale. This indicates that CICE has the potential to improve its seasonal forecast skill and provide more accurate predictions of sea ice extent and thickness when initialized with a more realistic sea ice thickness. This study highlights that the suitability of CICE for seasonal prediction depends on various factors, including initial conditions such as sea ice thickness, in addition to sea ice coverage, as well as oceanic and atmospheric conditions.

A novel seasonal grey prediction model with time-lag and interactive effects for forecasting the photovoltaic power generation

Electricity is one of the most important energy sources for the progress of human society. Photovoltaic power generation (PPG) has great potential as a renewable energy source to support sustainable development. However, the unpredictability of photovoltaics because of the uncertain weather conditions, time lags and the complex interactions among related factors may result in instabilities in PPG, which limits its applications in the electric power system. Therefore, accurately identifying the related factors and representing the interactions are the key issues in predicting PPG. For this purpose, a novel seasonal grey prediction model with time-lag and interactive effects, denoted as SGMLI(1,n), is developed to predict the capacity of photovoltaic power generation. More specifically, a time-lag driving term, a periodic driving term, and an interaction driving term are proposed to represent the lag relationship, periodicity, and the interactive effects between two factors, respectively. Utilizing the least square method and the Whale Optimization Algorithm (WOA), the optimal solution to the nonlinear programming problem is estimated. For elaboration and verification, six grey forecasting models, two statistical techniques, and two machine learning models, are employed for comparison with the SGMLI(1,n) model. Experimental results reveal that the SGMLI(1,n) outperforms benchmark models in forecasting PPG.

Can seasonal prediction models capture the Arctic mid‐latitude teleconnection on monthly time scales?

AbstractThis study explores Arctic warming's effect on Eurasia's temperature variability, notably the warm Arctic–cold Eurasia (WACE) pattern, and assesses seasonal prediction models' accuracy in capturing this phenomenon and its monthly variation. Arctic warming events are categorized into deep Arctic warming (DAW), shallow Arctic warming (SAW), warming aloft (WA), and no Arctic warming (NOAW), based on the temperatures at 2 m and 500 hPa in the Barents–Kara Sea. It is revealed that DAW events are significantly correlated with monthly cold temperature anomalies in East Asia, predominantly occurring in January–February, excluding December. This study evaluates two primary capabilities of seasonal prediction models: their proficiency in forecasting these Arctic warming events, particularly DAW, and their ability to replicate the spatial patterns associated with DAW. Some models demonstrated notable predictive skill for DAW events, with enhanced performance in January and February. Regarding spatial pattern reproduction, models showed limited alignment with the reference dataset over the Northern Hemisphere (above 25° N) in December, whereas a higher degree of concordance was observed in January–February. This indicates their capability in capturing the atmospheric circulation patterns associated with DAW, pointing to areas where model performance can be enhanced.

Research on Pricing and Replenishment of Vegetables Based on Grey Relational Model and Seasonal Index Prediction Model

This article mainly studies the problem of automatic pricing and replenishment decisions for vegetable products in supermarkets. A grey correlation model and ARIMA time series model are comprehensively established, and based on clustering analysis and regression analysis methods, the distribution patterns of sales volume for each vegetable category and item over time, as well as the relationship with cost plus pricing and mutual correlation, are analyzed to make reasonable predictions for future sales volume. In order to make decisions on product pricing, sales combination, and replenishment, and maximize supermarket profits, the data is preprocessed through normal test and quartile method. Through screening and sorting analysis, the distribution patterns of sales volume for each vegetable category and item over time are obtained; In order to analyze the relationship between different categories of vegetables and between different single products, a grey correlation model is established to determine the analysis sequence, and the correlation coefficients between each category and each single product are calculated to analyze the mutual relationship. And combined with the hierarchical clustering analysis method, the vegetable categories were divided into three sales types: high, medium, and low, expanding the relationship between categories. At the same time, a periodic chart was drawn to determine the periodicity of sales volume distribution in each category. Then, statistical analysis was conducted on the data to determine the linear correlation between sales volume and cost plus pricing. Based on this, a linear regression model was established to analyze the linear relationship between sales volume and cost plus pricing for the six categories; In order to provide the total daily replenishment amount and pricing strategy for the next week, this article establishes a seasonal index prediction model for sales forecasting. Combining the seasonal distribution of each category, the seasonal index is used to predict the sales volume for the next week. Considering the goal of maximizing profits, the target function is listed to comprehensively analyze the replenishment and pricing strategy for the next week.

Automatic Pricing and Replenishment Decision-Making for Vegetable Commodities Based on Seasonal Index Prediction Models

This study provides an in-depth analysis of the relationship between sales, pricing, and product categories to construct a predictive model designed to optimize replenishment decisions and maximize superstore profits. The study begins with data preprocessing, integrating data coded according to individual products, on which the model is built and solved. In the first part, descriptive statistical analysis and visualization revealed that the sales volume of each category showed a cluster-like distribution, which was affected by seasonal variations. Based on the time series smoothness test, it is initially judged that each individual product is a smooth series, and the smoothness of the series is verified by ADF test. In the second part, the correlation coefficients of the sales volume of each category and single product were calculated by using Spearman correlation analysis, and the results were presented through heat maps, which revealed the strong correlation between different categories. For the relational equation between total sales volume and cost-plus pricing, an autoregressive and least squares model was developed, with cost-plus pricing as the dependent variable and total sales volume as the independent variable. For the seasonal index prediction model, the total daily replenishment and pricing at future dates were predicted by calculating the seasonal index. The results show that the replenishment volume of flowers and leaves fluctuates greatly, and it is recommended that pricing and replenishment volume be adjusted in a timely manner according to the actual sales situation; cauliflower, aquatic roots and tubers, and eggplant are relatively stable, and they should be adjusted moderately according to the market demand; and chili peppers and edibles are on a downward trend, and it is recommended that replenishment volume be reduced in order to avoid the inventory backlog.

Trend and interannual variability of summer marine heatwaves in the tropical Indian ocean: Patterns, mixed layer heat budget, and seasonal prediction

Marine heatwaves (MHWs) are extreme sea surface temperature (SST) events in all ocean basins, with far-reaching impacts on marine ecosystems and socio-economy. The leading patterns, trend, and interannual variability of summer MHWs in the tropical Indian Ocean (TIO) are investigated in this study. The first empirical orthogonal function (EOF) mode of frequency of MHWs exhibits a monopole pattern over the entire basin. This mode is highly associated with the concurrent Indian Ocean Basin warming, indicating a remarkable trend over the past four decades. The linear trend in MHW properties largely relates to increased summer Indian Ocean mean SST. The second EOF mode exhibits a zonal dipole with the MHW numbers increasing in the west and decreasing in the east. On the interannual timescale, the first two EOF modes are remotely affected by antecedent and concurrent El Niño events, respectively. The ocean mixed layer budget is utilized for examining the formation of different summer MHW patterns. During the preceding spring, the surface heat flux is important for the development of MHWs, while the ocean advections play a secondary role in the South Indian Ocean for the MHW monopole. Once the SST anomaly rises in summer, the ocean advections play a dominant role in maintaining the SST. Last, we assess the prediction skill of summer TIO MHWs by performing a bilinear seasonal statistical prediction model. Our results suggest the frequency of summer MHWs in the TIO could be predicted one season in advance. This study has great implications for understanding and predicting ocean extreme events in the TIO.

A novel seasonal grey prediction model with fractional order accumulation for energy forecasting

To accurately predict sequence data with seasonal characteristics, we combine data restart technology and fractional order accumulation into a novel seasonal grey model (FSGM (1,1, α)). The particle swarm optimization (PSO) algorithm is used to solve the fractional order and background value coefficients of the model, and the effectiveness of FSGM (1,1, α) is verified using three cases. Finally, we use FSGM (1,1, α) to predict quarterly electricity generation in Beijing and Henan Province and quarterly petroleum coke production in China from 2023 to 2027. The research results indicate that, first, FSGM (1,1, α) is reasonable and effective and has the ability to accurately capture the dynamic trend of seasonal data. Second, compared with the grey model (GM (1,1)), seasonal grey model (SGM (1,1)), data grouping grey model (DGGM (1,1)), data grouping seasonal model (DGSM (1,1)), and data grouping seasonal time model (DGSTM (1,1)), which have seasonal characteristics, FSGM (1,1, α) can better fit the original data, achieve higher prediction accuracy, and perform better. Third, from 2023 to 2027, it is predicted that there will be no significant change in Beijing's electricity generation, and the current stable trend will be maintained. Both the power generation in Henan Province and the petroleum coke production in China will steadily increase to a certain extent, with obvious seasonal cyclical fluctuations. Notably, the power generation and petroleum coke production in Henan Province in the fourth quarter of 2027 will increase by 11.50 % and 10.93 %, respectively, compared to those in the fourth quarter of 2023.

A Simple Statistical Postprocessing Scheme for Enhancing the Skill of Seasonal SST Predictions in the Tropics

Abstract Seasonal prediction systems are subject to systematic errors, including those introduced during the initialization procedure, that may degrade the forecast skill. Here we use a novel statistical postprocessing correction scheme that is based on canonical correlation analysis (CCA) to relate errors in ocean temperature arising during initialization with errors in the predicted sea surface temperature fields at 1–12-month lead time. In addition, the scheme uses CCA of simultaneous SST fields from the prediction and corresponding observations to correct pattern errors. Finally, simple scaling is used to mitigate systematic location and phasing errors as a function of lead time and calendar month. Applying this scheme to an ensemble of seven seasonal prediction models suggests that moderate improvement of prediction skill is achievable in the tropical Atlantic and, to a lesser extent, in the tropical Pacific and Indian Ocean. The scheme possesses several adjustable parameters, including the number of CCA modes retained, and the regions of the left and right CCA patterns. These parameters are selected using a simple tuning procedure based on the average of four skill metrics. The results of the present study indicate that errors in ocean temperature fields due to imperfect initialization and SST variability errors can have a sizable negative impact on SST prediction skill. Further development of prediction systems may be able to remedy these impacts to some extent. Significance Statement The prediction of year-to-year climate variability patterns, such as El Niño, offers potential benefits to society by aiding mitigation and adaptation efforts. Current prediction systems, however, may still have substantial room for improvement due to systematic model errors and due to imperfect initialization of the oceanic state at the start of predictions. Here we develop a statistical correction scheme to improve prediction skill after forecasts have been completed. The scheme shows some moderate success in improving the skill for predicting El Niño and similar climate patterns in seven prediction systems. Our results not only indicate a potential for improving prediction skill after the fact but also point to the importance of improving the way prediction systems are initialized.

Effective Deep Learning Seasonal Prediction Model for Summer Drought Over China

AbstractDrought is an important meteorological event in China and can cause severe damage to both livelihoods and socio‐ecological systems, but current seasonal prediction skill for drought is far from successful. This study used convolutional neural network (CNN) to build an effective seasonal forecast model for the summer consecutive dry days (CDD) over China. The principal components (PC) of the six leading empirical orthogonal function modes of CDD anomaly were predicted by CNN, using the previous winter precipitation, 2‐m temperature and 500 hPa geopotential height as predictors. These predicted PCs were then projected onto the observed spatial patterns to obtain the final predicted summer CDD anomaly over China. In the independent hindcast period of 2007–2018, the interannual variabilities of first six PCs were significantly predicted by CNN. The spatial patterns of CDD were then skillfully predicted with anomaly correlation coefficient skills generally higher than 0.2. The interannual variability of summer CDD over the middle and lower Yangtze River valley, northwestern China and northern China were significant predicted by our CNN model three months in advance. CNN identified that the previous winter precipitation was the important predictor for PC1–PC3, whereas the previous winter 2‐m temperature and 500 hPa geopotential height were important for the prediction of PC4–PC6. This research provides a new and effective method for the seasonal prediction of summer drought.

Yellow River Water and Sand Monitoring Data Model based on SARIMAX Rediction

With the evolution of history, the problems faced by the Yellow River Basin have been an important research topic in hydrology. In this paper, for the problems in the monitoring of water and sand in the Yellow River, a seasonal time series model of the monitoring data is established based on the time series prediction model, and the CUSUM algorithm is used to test the model. First, the data are processed to calculate the spearman correlation for each pair of categories. Then, for analyzing the description of single dimensional data, data descriptive analysis was used. Next, the data were forecasted using a seasonal time series forecasting model. Finally, the seasonal time series prediction model was used to predict water fluxes and solved using double interpolation to obtain the predicted effect plots for the next ten years. The proposed model is close to the reality, can reasonably solve the proposed problem, and is characterized by high practicality and high efficiency of the algorithm.

The strengthened role of new predictors of Indian Ocean Dipole (IOD) during the recent decades of weakened ENSO-IOD relationship

Indian Ocean Dipole (IOD) is the major interannual ocean-atmosphere interaction phenomena in the Indian Ocean (IO) and is influenced by external forcing such as El Nino Southern Oscillation (ENSO). Meanwhile, its co-occurrence and stronger relationship with ENSO have decreased during recent decades. The IOD variability also reduced after 2000, accompanied by a shift of the western pole to central IO extending further southward. The study reports that the boreal fall (SON, September, October, November) season IOD has an intensified relationship with the previous winter Subtropical Indian Ocean dipole (SIOD) and spring season equatorial north tropical Atlantic (NTA) SST anomalies., while the ENSO relationship is reduced from its pre-2000 value. It is found that the persistent warming (cooling) in the western side of positive (negative) SIOD during the previous winter induces easterly (westerly) wind anomalies in the equatorial IO during the following summer and fall, leading to positive (negative) IOD events. These IOD events have the western pole shifted to south-central IO instead of the canonical northwestern warming. The spring season NTA SST anomalies induce stronger summer season circulation and SST gradient in the equatorial Pacific, like ENSO. During the SON season, this pattern is associated with cooling and easterly wind anomalies in the tropical eastern IO and IOD. These two patterns explain the major mode of IOD variability after 2000. While the IOD associated with NTA have co-occurring ENSO during summer and fall, the SIOD-induced IOD events are independent of ENSO in the Pacific. Thus, these two predictors provide long-lead predictability (2–3 seasons ahead) of IOD for both ENSO co-occurring and non-ENSO IOD events. However, the IOD predictability of seasonal prediction models mainly depends on the ENSO-IOD relationship, resulting in reduced IOD skills for many of them after 2000. The models such as COLA-CCSM4, which has improved skill have stronger than observed ENSO-IOD relationship than the pre-2000 period. A linear regression model including SIOD and NTA SST indices of the previous winter and spring season respectively as predictors simulates IOD with a skill of around 0.65 during the recent period, indicating the necessity of seasonal prediction models to capture the variability in the southern IO and NTA and their teleconnections for better prediction of IOD in the recent period of reduced ENSO skill.

A new hybrid model SARIMA-ETS-SVR for seasonal influenza incidence prediction in mainland China.

Seasonal influenza is a serious public health issue in China. This study aimed to develop a new hybrid model for seasonal influenza incidence prediction and provide reference information for early warning management before outbreaks. Data on the monthly incidence of seasonal influenza between 2004 and 2018 were obtained from the China Public Health Science Data Center website. A single seasonal autoregressive integrated moving average (SARIMA) model and a single error trend and seasonality (ETS) model were built. On this basis, we constructed SARIMA, ETS, and support vector regression (SARIMA-ETS-SVR) hybrid model. The prediction performance was determined by comparing mean absolute error (MAE), mean square error (MSE), mean absolute percentage error (MAPE), and root mean square error (RMSE) indices. The optimum SARIMA model was SARIMA (0,1,0) (0,0,1)12. Error trend and seasonality (ETS) (M,A,M) was the SARIMA optimal model. For the fitting performance, the SARIMA-ETS-SVR hybrid model achieved the lowest values of MAE, MSE, and RMSE, in addition to the MAPE. In terms of predictive performance, the SARIMA-ETS-SVR hybrid model had the lowest MAE, MSE, MAPE, and RMSE values among the three models. The study demonstrated that the SARIMA-ETS-SVR hybrid model provides better generalization ability than a single SARIMA model and a single ETS model, and the predictions will provide a useful tool for preventing this infectious disease.

Seasonal Forecasting of Precipitation, Temperature, and Snow Mass over the Western United States by Combining Ensemble Postprocessing with Empirical Ocean–Atmosphere Teleconnections

Abstract Accurate and reliable seasonal forecasts are important for water and energy supply management. Recognizing the important role of snow water equivalent (SWE) for water management, here we include the seasonal forecast of SWE in addition to precipitation (P) and 2-m temperature (T2m) over hydrologically defined regions of the western United States. A two-stage process is applied to seasonal predictions from two models (NCEP CFSv2 and ECMWF SEAS5) through 1) postprocessing to remove biases in the mean, variance, and ensemble spread and 2) further reducing the residual errors by linear regression using climate indices. The adjusted forecasts from the two models are combined to form a superensemble using weights based on their prior skill. The adjusted forecasts are consistently improved over raw model forecasts probabilistically for all variables and deterministically for SWE forecasts. Overall skill of the superensemble usually improves upon the skill of forecasts from individual models; however, the percentage of seasons and regions with increased skill was approximately the same as those with decreased skill relative to the top performing postprocessed individual model. Seasonal SWE has the highest prediction skill, followed by T2m, with P showing lower prediction skill. Persistence contributes strongly to the skill of SWE and moderately to the skill of T2m. Furthermore, a distinct seasonality in the skill is seen in SWE, with a higher skill from late spring through early summer. Significance Statement Here we test the postprocessing of seasonal forecasts from two state-of-the-art seasonal prediction models for traditionally forecasted elements of precipitation and temperature as well as snowpack, which is important for water management. A two-stage procedure is utilized, including ocean and atmospheric teleconnection indices that have been shown to impact seasonal weather across the western United States. First, we adjust model output based on the average error in historic runs and then relate the remaining error to these teleconnection indices. A final step combines each adjusted model based on its historic performance. Forecasts are shown to improve upon the original models when assessed probabilistically. The snowpack forecasts perform better than temperature and precipitation forecasts with the best performance from late winter through early summer. Persistence is found to contribute strongly to the skill of snowpack and moderately to the skill of temperature.

Seasonal precipitation forecasting for water management in the Kosi Basin, India using large-scale climate predictors

AbstractA novel approach for qualitative seasonal forecast of precipitation at a basin scale is presented as significant enhancement in seasonal forecast at regional and country scales in India. The process utilizes empirical and typically lagged relationships between target variables of interest, namely precipitation at the basin level and various large-scale climate predictors (LSCPs). A total of 14 LSCPs have been considered for the seasonal forecast of precipitation with lead times of 1, 2, and 3 months in the Kosi Basin, India. Random split training and testing were conducted on seven machine-learning (ML) models using a potential predictor dataset for model selection. The Logistic Regression (LR) model was adopted since it had the highest mean accuracy score compared to the remaining six ML models. The LR model has been optimized by testing it on all possible combinations of potential predictors using Leave-One-Out Cross-Validation (CV) scheme. The resulting Seasonal Prediction Model (SPM) provides the probability of each tercile categorized as Above Normal (AN), Normal (N), and Below Normal (BN). The model has been evaluated using various metrics.

Different role of spring season Atlantic SST anomalies in indian summer monsoon rainfall (ISMR) variability before and after early 2000

The present study shows that the weak (out-of-phase) correlation between the boreal spring Atlantic Meridional Mode (AMM) index and Indian summer monsoon rainfall (ISMR) during the 1990–2001 periods turned into a significant positive relationship after that. During the first period, boreal spring (MAM) season Atlantic SST variability is dominated by stronger cooling in the southeastern region. After 2002, the spring season SST pattern has north Atlantic warming extending to the equatorial region and continuing to summer (JJAS). The primary mode of the Atlantic SST co-occurred with a pre-existing El Nino during period 1 and the formation of La Nina by summer during period 2. During period 1, ENSO-induced anomalies dominated with weak divergence over India and the Atlantic region, while during period 2, the warm SST anomalies of north Atlantic-induced off-equatorial convection and associated circulation extended to the Indian Ocean and Indian land region resulting in increased monsoon rainfall during JJAS. The study further showed that during period 2, the increased correlation of ENSO and ISMR is also contributed by the Atlantic SST anomalies with additional off-equatorial wind anomalies and circulation from the north Atlantic extending to the Indian Ocean and monsoon region. Thus, MAM season Atlantic SST anomalies play a significant role in ENSO phase reversal and ISMR during the recent period. Many of the seasonal prediction models that participated in the NMME project capture the phase reversal of AMM-ISMR correlation when initialized during February. But models have more robust equatorial SST patterns during the summer season, and the north Atlantic SST anomalies are vital in the absence of ENSO. The study indicates that the spring season AMM index can provide a predictive signal for ISMR in seasonal prediction models. Still, they need to improve the simulation of the Atlantic basic state and its teleconnection with the tropical east Pacific.

Multi-model seasonal prediction of global surface temperature based on partial regression correction method

The increased climate change is having a huge impact on the world, with the climatic change sensitive and vulnerable regions at significant risk particularly. Effective understanding and integration of climate information are essential. It helps to reduce the risks associated with adverse weather conditions and to better adapt to the impacts of climate variability and change. Using the hindcast data from Japan Meteorological Agency/Meteorological Research Institute (JMA/MRI) coupled prediction system version 2 (JMA/MRI-CPS2), National Centers for Environmental Prediction (NCEP) Climate Forecast System model version 2 (CFSv2), and Canadian Centre for Climate Modeling and Analysis (CCCma) Coupled Climate Model, versions 3 (CanCM3) seasonal prediction model systems, the performance of seasonal prediction for global surface temperature in boreal summer and winter is comprehensively evaluated and compared for 1982–2011 from the perspective of deterministic and probabilistic forecast skills in this study, and a partial regression correction (PRC) method is introduced to correct seasonal predictions. The results show high prediction skills in the tropics, particularly in the equatorial Pacific, while poor skills on land. In general, JMA/MRI-CPS2 has slightly better prediction performance than CFSv2 and CanCM3 in the tropics. CFSv2 is generally superior to JMA/MRI-CPS2 and CanCM3 in the extratropical northern hemisphere and East Asia, especially for the abnormal low winter temperature prediction in East Asia. CanCM3 shows good deterministic forecast skills in extra-tropics but performs slightly worse in probabilistic forecasting. Based on the respective strengths of each seasonal prediction model, an ensemble forecast correction with observational constraint is implemented by partial regression, and the improvement of skills in ensemble predicting has been analyzed. Compared to the simple multi-model ensemble (MME), the correction improved the global-average temporal correlation coefficient and multi-year mean anomaly correlation coefficient by about 0.1 and 0.13, respectively. The validation tests indicate that the corrected ensemble forecast has higher ranked probability skill scores than that of the MME, which is improved by more than 0.06 in the tropics. Meanwhile, when the training period is sufficiently long, it may have the potential for future seasonal temperature predictions from the perspective of stable zonal partial regression coefficients.

A skilful seasonal prediction for wintertime rainfall in southern Thailand

AbstractYear‐to‐year variations of southern Thailand rainfall (STR) in boreal winter exert profound social and economic impacts, whereas current multimodel ensemble prediction systems have low skills to predict the STR. This study proposes a physical‐based seasonal prediction model for the winter STR 1 month in advance using the outputs from the dynamic models. The prediction model is constructed using linear regression, with the tropical western Pacific (TWP) sea surface temperature (SST) anomaly in preceding October as a predictor. Its prediction skill in the leave‐five‐out cross‐validation is significantly higher than that of the multimodel ensemble mean. The mechanism behind this model is also discussed. In October, the warm TWP SST anomalies can trigger anomalous low‐level convergence surrounding the South China Sea in terms of the Matsuno–Gill mechanism and persist into the following winter, causing above‐than‐normal STR. This information is essential and may provide another perspective to improve the model prediction on the winter STR.