Teleconnection events can influence normal regional weather patterns and affect weather forecast accuracy. To improve the forecast ability, the relationship between main teleconnections such as El Niño–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Madden–Julian Oscillation (MJO), and climate variables (rainfall, maximum and minimum surface temperature, vertical mixing ratio, and vertical maximum temperature) was established using lag correlation coefficient and t-test methods. The results reveal moderately significant correlations between El Niño, positive IOD and rainfall, and vertical mixing ratio, which can be associated with lower-than-usual rainfall. The coincidence between El Niño and positive IOD events can worsen drought. Even though the MJO and regional weather correlations were significant, the magnitude of correlation coefficients was negligible. In addition, the spatiotemporal distribution of ENSO shows that the strong El Niño has more influence on rainfall anomalies in the post-1980s. Since there are insufficient studies on the association between teleconnections and climate variables, especially vertical mixing ratio, our findings can benefit prediction development for teleconnection-induced regional climate anomalies for extreme events and water management preparations in northern and northeastern Thailand.

  • ENSO events have significant correlations in northern compared with northeastern Thailand.

  • El Niño, positive IOD, and climate variables have significant negative correlations in the dry season in northern Thailand.

  • La Niña, negative IOD, and climate variables have a less significant negative correlation in northern and northeastern Thailand.

  • MJO has significant correlations with rainfall and temperature.

Teleconnection patterns are defined as recurring and persistent large-scale patterns of pressure and circulation anomalies that can influence long-distance regional weather patterns far from the source through wind circulations. The implication of weather patterns varies with regions and seasons (NOAA 2020; Clarke et al. 2022; Zavadoff & Arcodia 2022). The well-known teleconnections such as El Niño–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), and Madden–Julian Oscillation (MJO) are mostly sea surface temperature (SST) anomalies in the tropical Pacific, Indian Ocean, and Tropical areas (Feldstein & Franzke 2017). These events can influence significant global effects on weather variations, especially the precipitation distribution (Rodríguez-Fonseca et al. 2016; Cash et al. 2017; Aamir & Hassan 2020; Williams et al. 2023). Furthermore, some worldwide extreme weather events are found to be associated with teleconnection phenomena (Räsänen & Kummu 2013; Czymzik et al. 2016; Boers et al. 2019; Ciemer et al. 2020; Mishra et al. 2022; Liu et al. 2023). Therefore, the relationship between teleconnections and various climate variables should be investigated to improve the reliability of weather prediction systems.

The relationship between teleconnections, precipitation, and surface temperature has been examined by many studies, which mostly confirm the linkage of the negative teleconnection phase with lower-than-usual rainfall and the positive teleconnection phase with higher-than-usual rainfall (Phillips et al. 2012; Irwandi et al. 2018; Sreekala et al. 2018; Li et al. 2021). Aamir et al. (2022) ascertained that in Pakistan, the influence of teleconnection indices on monthly precipitation patterns is weak in winter months and moderate to strong in monsoon month. Moreover, Lüdecke et al. (2021) studied the influence of potential teleconnections with continent-wide African rainfall by 11-month time lag correlation to improve short- to mid-term rainfall prognoses. Lee et al. (2023) found that the United States precipitation responds to ENSO events with a high significance level of an opposite tendency. In addition, the likelihood that heat waves and extreme precipitation can occur in two consecutive seasons is higher during El Niño than La Niña events (Mishra et al. 2022). Even though the concern of global temperature rise from anthropogenic climate change might increase the teleconnection events, there is overall low confidence about the future changes of magnitude, frequency, and spatial distribution of these patterns (IPCC 2021).

Thailand's weather patterns are influenced by prevailing monsoons, namely the Southwest Monsoon, which brings abundant warm moist airmass from the Indian Ocean resulting in the rainy season from mid-May to mid-October (World Bank Group 2021). The Northeast Monsoon brings cold, dry airmass from mainland China to Thailand during the dry season in upper Thailand, namely, central, northern, and northeastern. However, this monsoon brings moisture from the Thai Gulf to southern Thailand resulting in abundant rainfall from mid-October to mid-February (Thai Meteorological Department 2015). The strong link between ENSO and monsoons in Thailand has been found since 1980 and was linked to several extreme events (Singhrattna et al. 2005). For instance, from 1997 to 1998, El Niño year, the observed rainfall amounts data from the Thai Meteorological Department (TMD) indicated below-normal rainfall by 5.47% and 14.46%, respectively, whereas in La Niña year 1999, the rainfall amount in Thailand was above normal by 36.59% (Ueangsawat 2013; Promchote et al. 2016).

Several studies have attempted to find the relationship between major teleconnections and extreme events in Thailand for better weather predictions (Sahu et al. 2010; Wikarmpapraharn & Kositsakulchai 2010; Muangsong et al. 2014; Puttrawutichai et al. 2017). Moreover, the studies showed that rainfall associated with ENSO, IOD, and MJO results in different magnitudes and timing for correlation values in different parts of Thailand. However, some studies revealed that the significant relationship between ENSO, IOD, and rainfall patterns could not be concluded (Nounmusig 2018; Bridhikitti 2019). Besides, vertical humidity and temperature determine the probability of magnitude and intensity of rainfall, which can lead to droughts or heavy rainfall. Nevertheless, the study of the relationship between teleconnections and vertical humidity in Thailand is still insufficient. Trakolkul et al. (2022) investigated the moderate to severe correlation between ENSO and precipitable water vapor (PWV) from 11 stations during 2007–2016. La Niña events were highly correlated with PWV, especially in Thailand's severe flooding in 2011. Furthermore, the PWV declined at the end of 2011 and consequently caused flooding and severe storms.

The percentage of the agricultural area in northern and northeastern Thailand is 23.6% and 46.8% (NSO 2013). However, northern and northeastern Thailand is susceptible to extreme weather, such as floods and droughts, adversely affecting rainfed and irrigated agricultural systems (UNDP 2023). Moreover, 80% of Thailand's severe drought-prone areas occur in the north and northeast regions (World Bank Group 2021). Agriculture is strongly affected by weather and climate variability (Hoogenboom 2000). The crop yield is sensitive to teleconnection patterns; therefore advance notice to farmers for adaptation and mitigation of any potential losses can be provided with proper analysis of the teleconnection influences (Gonsamu & Chen 2015; Ceglar et al. 2017). UNICEF (2023) reported that Thai children are at high risk from climate change and environmental degradation such as extreme heat, cold, floods, and droughts, especially in northeast and southern Thailand. In addition, Chiang Mai and Chiang Rai provinces in northern Thailand were ranked top ten in floods and cold risks. Furthermore, northern and northeastern areas may face riverine flooding during the rainy season (Lim & Boochabun 2012) and flash floods in mountainous areas or the foothills (Thai Meteorological Department 2015).

This study aims to determine the relationship patterns between major teleconnections and important climate variables that are crucial to extreme weather predictions. Furthermore, the implication of ENSO spatially and temporally was also investigated in the north and northeastern Thailand. The findings from this study could enhance our knowledge about the teleconnection implication of weather variation in northern and northeastern Thailand. In addition, the improvement of extreme weather prediction can be beneficial for disaster preparation and water resource management in northern and northeastern Thailand, which are vulnerable to various climatic disasters.

The approach for relationship pattern establishment between major teleconnections and climate variables is based on statistical lag correlation to measure the paired variables’ association strength and direction. Subsequently, the statistical significance is confirmed with the t-test method. The data and methodology used in the analysis are described in the following sub-sections.

Study area

Northern Thailand is located at latitude 15.3° to 20°N and longitude 97.8° to 102°E with topography characterized by north–south high mountain ranges (Figure 1). This area consists of 15 provinces that cover 153,000 km2 based on meteorological characteristics. The average annual rainfall during 1951–2021 is 1,299 mm, with approximately 88.6% of rainfall occurring in the rainy season during May–October. The annual average maximum temperature is 32.9 °C. The highest average maximum temperature month is April (37.2 °C). The annual average minimum temperature is 20.9 °C, and the lowest average maximum temperature month is January (15.1 °C).
Figure 1

An overview map of the study area. The inset map shows the location of the study area within Thailand and neighboring countries. A total of 21 selected TMD weather stations are represented by numbers and denote by blue dots for the northern area and red dots for the northeastern area. Blue stars denote two TMD upper-air measurement stations. The detail of locations and station names are shown in the Supplementary Material, Table S1.

Figure 1

An overview map of the study area. The inset map shows the location of the study area within Thailand and neighboring countries. A total of 21 selected TMD weather stations are represented by numbers and denote by blue dots for the northern area and red dots for the northeastern area. Blue stars denote two TMD upper-air measurement stations. The detail of locations and station names are shown in the Supplementary Material, Table S1.

Close modal

Northeastern Thailand's location is at latitude 14.13° to 18.45°N and longitude 100.9°E to 105.61°E (Figure 1). The dominant topography is a high plain with high mountain borders. This area consists of 20 provinces covering around 170,000 km2. The average annual rainfall is 1,480 mm, with approximately 88.7% of rainfall occurring in the rainy season. The annual average maximum temperature is 32 °C, and the highest average maximum temperature month is April (35.9 °C). The annual average minimum temperature is 21.6 °C, and the lowest average maximum temperature month is January (16.2 °C).

The main source of agricultural water in north and northeastern Thailand is monsoon rainfall. During the rainy season, a ‘dry spell’ may occur around the end of June to mid-July, resulting in low rainfall (Thai Meteorological Department 2015).

Data collection

The relationship patterns between teleconnections and climate variables were established using the selected teleconnection indices and main climate variables. Detail about the data collected is explained in Table 1 and subsequent sections.

Table 1

Sources and types of data used in this study

No.DataResolutionPeriodSource
I. Teleconnection indices   
 ONI (ENSO) 3-MRM 1950–2021 NOAA 
 DMI (IOD) Monthly 1870–2021 NOAA 
 RMM (MJO) Daily 1974–2021 BOM 
II. Surface climate variables   
 Rainfall Point/Daily 1951–2021 TMD, RID 
 Maximum temperature    
 Minimum temperature Point/Daily 1951–2021 TMD 
III. Vertical climate variables   
 Upper-air data Point/Daily 1973–2021 UW (TMD) 
 Vertical temperature    
 Vertical dew point temperature    
No.DataResolutionPeriodSource
I. Teleconnection indices   
 ONI (ENSO) 3-MRM 1950–2021 NOAA 
 DMI (IOD) Monthly 1870–2021 NOAA 
 RMM (MJO) Daily 1974–2021 BOM 
II. Surface climate variables   
 Rainfall Point/Daily 1951–2021 TMD, RID 
 Maximum temperature    
 Minimum temperature Point/Daily 1951–2021 TMD 
III. Vertical climate variables   
 Upper-air data Point/Daily 1973–2021 UW (TMD) 
 Vertical temperature    
 Vertical dew point temperature    

Note: ONI, Oceanic Niño Index; ENSO, El Niño–Southern Oscillation; DMI, Dipole Mode Index; IOD, Indian Ocean Dipole; RMM, Real-time Multivariate MJO Index; MJO, Madden–Julian Oscillation; NOAA, National Oceanic and Atmospheric Administration; BOM, Bureau of Meteorology; TMD, Thai Meteorological Department; RID, Royal Irrigation Department; UW, University of Wyoming; 3-MRM, three-month running mean.

Surface climate variables

Observed daily rainfall, and the maximum and minimum surface temperature during 1951–2021 are acquired from TMD and Royal Irrigation Department (RID) weather stations. A total of 21 TMD weather stations distributed with daily data ≥ 30 years and missing data ≤ 10% across the study region were selected. The nearby station method filled in the missing data using the daily data from RID stations. The detail of paired TMD and RID stations are depicted in the Supplementary Material, Table S2.

Vertical climate variables

The observed daily data at Coordinated Universal Time (UTC) 0000Z during 1973–2021 were acquired from two TMD upper-air measurement stations, namely Chiangmai station (VTCC) in northern and Ubon Ratchathani station (VTUU) in northeastern Thailand. Moreover, the daily vertical dew point temperature and temperature at the surface, 850, 700, 500, and 300 hPa, which are standard levels, were used for the water vapor mixing ratio for representing vertical humidity and temperature calculations (MetOffice 2011). The upper-air measurement data used in this study were retrieved from the University of Wyoming website (https://weather.uwyo.edu/upperair/).

Teleconnection indices

Teleconnection indices are the ocean–atmosphere interaction events in one place that can influence weather patterns in a distant part of the world (Zavadoff & Arcodia 2022). The teleconnections selected for analysis in this study are ENSO, IOD, and MJO. For ENSO, the ONI which is NOAA's primary indicator for monitoring ENSO states was used; the ONI values are positive (negative) during El Niño (La Niña) phases (NOAA 2020). The second teleconnection index is the Dipole Mode Index (DMI), commonly used to identify IOD. IOD events may start in May or June, mature during August and October, and rapidly decay until around January (Bureau of Meteorology 2021a). The Real-time Multivariate MJO Index (RMM MJO) is the index for identifying MJO status. Overall, eight phases of MJO also affect opposite convective (suppressive) phases of the dipole (NOAA 2020).

ONI in 3-mrm values was used to find the relationship between ENSO, IOD, and climate variables in each month. Moreover, monthly DMI (IOD) was also used to find the association with monthly climate variables, so the association of both teleconnections to monthly climate variables could be compared. However, the MJO cycle is shorter than ENSO and IOD, therefore the relationship between MJO and climate variables was calculated weekly.

Overall methodology

The relationship association between teleconnections and climate variables and ENSO influence on spatial and temporal rainfall in northern and northeastern Thailand were assessed using different statistics. Since teleconnections might influence climate variables on different timescales, based on previous studies (Sattar et al. 2021; Wijeratne et al. 2022), the methods in this study are separated into two sections. Firstly, the lag correlation was applied to investigate teleconnections' influence on monthly and weekly climate variables. Secondly, the warm and cold phases of ENSO events' influence on annual rainfall were visualized in spatial and temporal maps.

Teleconnections vs climate variable association

Two statistical methods were employed to establish the relationship between teleconnection indices and climate variables. First, the 0–11 monthly lag correlation coefficient method was applied to find the lag and influence of each pair of variables. Subsequently, the t-test method was applied to confirm the statistical significance at 95% (p-value < 0.05). The pair variables are as shown in Figure 2.
Figure 2

The methodological flowchart depicts statistical methods used to establish relationship patterns between teleconnection indices vs climate variables.

Figure 2

The methodological flowchart depicts statistical methods used to establish relationship patterns between teleconnection indices vs climate variables.

Close modal

The daily rainfall from weather stations was summarized into weekly and monthly data – subsequently, the average between northern and northeastern stations. Maximum and minimum temperatures were averaged into monthly and weekly data. The 95th percentile values for monthly rainfall, maximum temperature, and the fifth percentile for monthly minimum temperature were calculated to assess extreme conditions. Subsequently, the monthly and weekly climate variables were used to perform lag correlation and t-tests with negative and positive phases of teleconnection indices. The 3-mrm ONI (ENSO) performed lag correlation with monthly climate variables, and the first or zero lag starts with the first month of ONI; for instance, the ONI index in DJF month will correlate lag 0 with December rainfall, lag 1 with January rainfall and so on until lag 11 follows (Ueangsawat 2013). Furthermore, IOD resolution is monthly, therefore, the lag correlation was performed with monthly climate variables from 0 to 11 lags as well. Moreover, MJO vs climate variables were correlated by weekly lag correlation.

The daily upper-air measurement data at the levels of surface, 850, 700, 500, and 300 hPa was used to calculate the monthly water vapor mixing ratio and average monthly temperature. To find the individual effect of teleconnection to humidity, the water vapor mixing ratio, which is invariant to temperature, was applied in this study as per Equation (1) (AMS 2022):
(1)
where r is the water vapor mixing ratio, e is the vapor pressure, ρ is the pressure.
The correlation coefficient between variables was performed using the non-parametric Spearman's rank correlation coefficient method. Spearman's rank correlation coefficient can determine the strength and direction of the monotonic relationship and robustness to outliers (Zaiontz 2022). Spearman's rank correlation coefficient without tied rank is described in Equation (2):
(2)
where di = rank xi – rank yi,ρ is the Spearman's rank correlation coefficient, n is the number of observations.

In this study, the correlation coefficient and t-test calculation were done using the Real Statistics Resource Pack for Excel free software (Zaiontz 2022). The positive and negative values of teleconnection indices and climate variable values can affect the direction of correlation. Therefore, the interpretation of correlation directions of the results between positive (negative) indices and positive (negative) variables is summarized in the Supplementary Material, Table S3.

Multiple testing

The multiple testing problem can occur in collective significance testing for determining significant patterns or an overall effect. This problem can lead to spurious significance results (Cortés et al. 2020). In this study, the statistical t-test was performed with an 11-lag correlation for each teleconnection index and climate variable, therefore our results may suffer from false-positive errors. The correction method such as the familywise error rate of Bonferroni can increase false-negative errors, due to very low adjusted p-values. The objective of this study is to explore the connection between teleconnections and regional climate variables in north and northeastern Thailand, therefore the correction of p-values is not employed. Hence, detailed confirmation of the present results may be required in future research (Lüdecke et al. 2021).

The spatiotemporal influence of ENSO on annual rainfall

The spatiotemporal influence of ENSO on north and northeast Thailand was studied using the Inverse Distance Weighted (IDW) method to construct rainfall anomaly maps from 1951–1980 and 1981–2021 (Chen & Liu 2012). Singhrattna et al. (2005) found a linkage between ENSO and Thailand monsoons was more closely related after 1980, which is contrary to the ENSO–Indian monsoon relationship. The ENSO indices are measured from SST anomalies, therefore the global temperature increase may affect the intensities of ENSO events. Hence, anomaly rainfall maps were constructed for 1951–1980 and 1981–2021 to investigate the temporal intensity between ENSO and rainfall. In addition, the influence of ENSO was considered by onset or ENSO start year and decay or ENSO decline year. Furthermore, the ENSO year intensities were classified into three groups based on the magnitude values of the ONI index. The threshold of each intensity is Strong ≥ 1.5, Moderate 1.0–1.4, and Weak 0.5–0.9; additional information on the ENSO state year is provided in the Supplementary Material, Table S4 (Golden Gate Weather Services 2022).

To investigate the implications of ENSO on different elevations, the weather stations were classified into two groups based on location heights. The threshold 200 m mean-sea-level (MSL) was chosen based on the height of the Khorat plateau, which covers most of northeastern Thailand. The weather stations at 200 m MSL or higher represent the highlands and mountainous areas, and those lower than 200 m MSL represent plain areas. Subsequently, the rainfall anomalies of ENSO events were visualized as contour maps.

The correlation coefficient (CC) results and its corresponding lag months are presented in Tables 27 and Figures 310. The highest CC values between teleconnection indices (TI) and climate variables (CV) are presented with a number of lag months in the bracket. The bold number indicates a significance level of 95% (p-value < 0.05); it is to be noted that the p-value is not adjusted (see Section 2.3.2). Details of how to read the tables and figures can be found in the Supplementary Material, Tables S5 and S6.
Table 2

Spearman's rank correlation coefficient values (ρ) for ENSO vs rainfall in northern and northeastern Thailand

DJFJFMFMAMAMAMJMJJJJAJASASOSONONDNDJ
El Niño vs Monthly Rainfall 
North             
R_M − 0.49(4) − 0.54(3) − 0.52(1) − 0.43(1) − 0.48(1) − 0.28(1) − 0.73(0) − 0.35(2) 0.36(10) − 0.27(7) − 0.43(11) − 0.46(5) 
R_95 − 0.47(4) − 0.44(3) − 0.64(6) 0.40(4) − 0.29(9) − 0.36(1) − 0.71(0) − 0.47(2) − 0.30(6) − 0.43(11) − 0.50(11) − 0.55(10) 
Northeast            
R_M − 0.41(3) 0.42(7) 0.49(4) − 0.75(0) − 0.44(1) 0.37(9) 0.36(8) − 0.41(2) 0.42(10) 0.53(1) 0.42(0) − 0.42(10) 
R_95 0.46(6) − 0.50(3) 0.63(4) − 0.75(0) − 0.57(5) 0.47(9) 0.39(4) 0.38(3) 0.52(10) 0.51(1) 0.48(0) − 0.45(10) 
La Niña vs Monthly Rainfall 
North             
R_M − 0.39(9) − 0.28(3) − 0.60(8) 0.65(4) 0.62(5) − 0.60(3) 0.43(3) − 0.33(5) − 0.41(8) − 0.48(7) 0.42(8) 0.38(7) 
R_95 − 0.34(5) 0.40(7) 0.49(7) 0.43(4) 0.66(5) − 0.51(3) 0.60(3) 0.40(2) 0.42(1) 0.54(0) 0.32(5) 0.33(7) 
Northeast            
R_M − 0.31(4) − 0.36(3) − 0.33(4) − 0.37(11) 0.49(5) 0.33(4) 0.44(2) 0.33(8) − 0.36(11) 0.55(9) 0.33(8) − 0.41(5) 
R_95 − 0.27(7) − 0.32(6) − 0.32(3) − 0.36(11) 0.45(5) − 0.32(8) 0.37(2) − 0.45(5) − 0.22(11) 0.49(9) 0.26(1) − 0.31(5) 
DJFJFMFMAMAMAMJMJJJJAJASASOSONONDNDJ
El Niño vs Monthly Rainfall 
North             
R_M − 0.49(4) − 0.54(3) − 0.52(1) − 0.43(1) − 0.48(1) − 0.28(1) − 0.73(0) − 0.35(2) 0.36(10) − 0.27(7) − 0.43(11) − 0.46(5) 
R_95 − 0.47(4) − 0.44(3) − 0.64(6) 0.40(4) − 0.29(9) − 0.36(1) − 0.71(0) − 0.47(2) − 0.30(6) − 0.43(11) − 0.50(11) − 0.55(10) 
Northeast            
R_M − 0.41(3) 0.42(7) 0.49(4) − 0.75(0) − 0.44(1) 0.37(9) 0.36(8) − 0.41(2) 0.42(10) 0.53(1) 0.42(0) − 0.42(10) 
R_95 0.46(6) − 0.50(3) 0.63(4) − 0.75(0) − 0.57(5) 0.47(9) 0.39(4) 0.38(3) 0.52(10) 0.51(1) 0.48(0) − 0.45(10) 
La Niña vs Monthly Rainfall 
North             
R_M − 0.39(9) − 0.28(3) − 0.60(8) 0.65(4) 0.62(5) − 0.60(3) 0.43(3) − 0.33(5) − 0.41(8) − 0.48(7) 0.42(8) 0.38(7) 
R_95 − 0.34(5) 0.40(7) 0.49(7) 0.43(4) 0.66(5) − 0.51(3) 0.60(3) 0.40(2) 0.42(1) 0.54(0) 0.32(5) 0.33(7) 
Northeast            
R_M − 0.31(4) − 0.36(3) − 0.33(4) − 0.37(11) 0.49(5) 0.33(4) 0.44(2) 0.33(8) − 0.36(11) 0.55(9) 0.33(8) − 0.41(5) 
R_95 − 0.27(7) − 0.32(6) − 0.32(3) − 0.36(11) 0.45(5) − 0.32(8) 0.37(2) − 0.45(5) − 0.22(11) 0.49(9) 0.26(1) − 0.31(5) 

Note: Numbers in () denote lag months, bold values indicate p-value < 0.05.

R_M indicates average monthly rainfall, R_95 indicates 95th percentile monthly rainfall. El Niño ONI values are positive (plus values) and La Niña ONI values are negative (minus values).

Table 3

Spearman's rank correlation coefficient values for IOD vs rainfall in northern and northeastern Thailand

JanFebMarAprMayJunJulAugSepOctNovDec
Positive IOD vs Monthly Rainfall 
North             
R_M 0.80(2) − 0.83(4) 0.95(5)  − 1.00(6) 0.81(2) − 0.68(7) − 0.75(8) − 0.71(8) − 0.55(11) − 0.71(11)  
R_95 0.80(3) − 0.77(4) 0.95(7)  0.80(2) − 0.68(7) − 0.77(7) − 0.49(8) − 0.47(8) − 0.47(9) − 0.70(11)  
Northeast            
R_M 1.00(10) − 0.54(8) 0.95(5)  1.00(5) − 0.74(7) − 0.71(7) − 0.63(8) − 0.65(2) − 0.66(6) − 0.48(1)  
R_95 1.00(10) 0.83(5) 0.95(5)  − 1.00(8) − 0.84(7) − 0.77(7) − 0.72(8) − 0.62(2) − 0.79(6) − 0.55(7)  
Negative IOD vs Monthly Rainfall 
North             
R_M  0.79(6) 0.80(0) − 0.89(4) 0.53(5) − 0.58(1) − 0.53(7) 0.81(8) 0.53(11) − 0.41(4) 0.94(2)  
R_95  0.75(0) 0.80(1) − 0.71(4) 0.51(5) 0.57(5) 0.49(5) 0.52(8) 0.41(9) − 0.58(4) 0.94(2)  
Northeast            
R_M  0.93(5) 0.70(2) − 0.77(4) − 0.55(10) 0.97(5) − 0.27(9) 0.72(8) 0.46(9) − 0.42(0) 0.89(2)  
R_95  0.86(5) 0.90(2) − 0.89(4) − 0.57(10) 0.97(5) 0.32(6) 0.45(8) 0.52(9) − 0.49(0) 0.84(2)  
JanFebMarAprMayJunJulAugSepOctNovDec
Positive IOD vs Monthly Rainfall 
North             
R_M 0.80(2) − 0.83(4) 0.95(5)  − 1.00(6) 0.81(2) − 0.68(7) − 0.75(8) − 0.71(8) − 0.55(11) − 0.71(11)  
R_95 0.80(3) − 0.77(4) 0.95(7)  0.80(2) − 0.68(7) − 0.77(7) − 0.49(8) − 0.47(8) − 0.47(9) − 0.70(11)  
Northeast            
R_M 1.00(10) − 0.54(8) 0.95(5)  1.00(5) − 0.74(7) − 0.71(7) − 0.63(8) − 0.65(2) − 0.66(6) − 0.48(1)  
R_95 1.00(10) 0.83(5) 0.95(5)  − 1.00(8) − 0.84(7) − 0.77(7) − 0.72(8) − 0.62(2) − 0.79(6) − 0.55(7)  
Negative IOD vs Monthly Rainfall 
North             
R_M  0.79(6) 0.80(0) − 0.89(4) 0.53(5) − 0.58(1) − 0.53(7) 0.81(8) 0.53(11) − 0.41(4) 0.94(2)  
R_95  0.75(0) 0.80(1) − 0.71(4) 0.51(5) 0.57(5) 0.49(5) 0.52(8) 0.41(9) − 0.58(4) 0.94(2)  
Northeast            
R_M  0.93(5) 0.70(2) − 0.77(4) − 0.55(10) 0.97(5) − 0.27(9) 0.72(8) 0.46(9) − 0.42(0) 0.89(2)  
R_95  0.86(5) 0.90(2) − 0.89(4) − 0.57(10) 0.97(5) 0.32(6) 0.45(8) 0.52(9) − 0.49(0) 0.84(2)  

Note: Numbers in () denote lag months, bold values indicate p-value < 0.05.

R_M indicates average monthly rainfall, R_95 indicates 95th percentile monthly rainfall. Positive IOD values are positive (plus values). Negative IOD values are negative (minus values).

Table 4

Spearman's rank correlation coefficient values for ENSO vs maximum surface temperature in northern and northeastern Thailand

DJFJFMFMAMAMAMJMJJJJAJASASOSONONDNDJ
El Niño vs maximum surface temperature 
North             
T_Max 0.68(4) 0.66(3) 0.57(6) 0.61(5) 0.33(2) 0.41(5) 0.59(9) 0.48(8) − 0.39(10) 0.29(7) 0.48(6) 0.59(5) 
T_95 0.56(4) 0.57(3) 0.63(1) 0.70(0) 0.57(4) 0.42(9) 0.51(9) 0.42(8) 0.40(0) 0.25(1) 0.34(6) 0.49(5) 
Northeast             
T_Max 0.53(4) 0.55(2) 0.50(2) 0.73(1) − 0.37(6) 0.43(10) 0.51(9) 0.50(4) 0.40(5) 0.28(10) 0.44(6) 0.49(5) 
T_95 0.47(3) 0.55(2) 0.62(3) 0.62(0) − 0.28(7) 0.59(10) 0.55(9) 0.46(8) 0.41(5) 0.39(6) 0.48(5) 0.47(4) 
La Niña vs maximum surface temperature 
North             
T_Max 0.44(5) 0.42(1) 0.46(1) 0.46(9) 0.46(2) 0.51(7) 0.48(10) 0.49(9) 0.61(8) − 0.53(10) 0.34(11) 0.38(6) 
T_95 0.36(4) 0.41(3) 0.43(9) 0.52(8) 0.46(7) − 0.42(9) 0.41(10) 0.42(11) 0.58(8) 0.57(7) 0.37(6) 0.40(5) 
Northeast             
T_Max 0.48(2) 0.73(1) 0.58(0) 0.45(3) 0.56(2) 0.49(7) 0.48(10) 0.50(3) 0.43(8) − 0.39(9) 0.54(0) 0.43(3) 
T_95 0.31(5) 0.46(3) 0.42(4) 0.44(8) 0.41(2) 0.42(3) − 0.60(3) 0.44(11) 0.46(9) 0.52(7) 0.44(7) 0.42(6) 
DJFJFMFMAMAMAMJMJJJJAJASASOSONONDNDJ
El Niño vs maximum surface temperature 
North             
T_Max 0.68(4) 0.66(3) 0.57(6) 0.61(5) 0.33(2) 0.41(5) 0.59(9) 0.48(8) − 0.39(10) 0.29(7) 0.48(6) 0.59(5) 
T_95 0.56(4) 0.57(3) 0.63(1) 0.70(0) 0.57(4) 0.42(9) 0.51(9) 0.42(8) 0.40(0) 0.25(1) 0.34(6) 0.49(5) 
Northeast             
T_Max 0.53(4) 0.55(2) 0.50(2) 0.73(1) − 0.37(6) 0.43(10) 0.51(9) 0.50(4) 0.40(5) 0.28(10) 0.44(6) 0.49(5) 
T_95 0.47(3) 0.55(2) 0.62(3) 0.62(0) − 0.28(7) 0.59(10) 0.55(9) 0.46(8) 0.41(5) 0.39(6) 0.48(5) 0.47(4) 
La Niña vs maximum surface temperature 
North             
T_Max 0.44(5) 0.42(1) 0.46(1) 0.46(9) 0.46(2) 0.51(7) 0.48(10) 0.49(9) 0.61(8) − 0.53(10) 0.34(11) 0.38(6) 
T_95 0.36(4) 0.41(3) 0.43(9) 0.52(8) 0.46(7) − 0.42(9) 0.41(10) 0.42(11) 0.58(8) 0.57(7) 0.37(6) 0.40(5) 
Northeast             
T_Max 0.48(2) 0.73(1) 0.58(0) 0.45(3) 0.56(2) 0.49(7) 0.48(10) 0.50(3) 0.43(8) − 0.39(9) 0.54(0) 0.43(3) 
T_95 0.31(5) 0.46(3) 0.42(4) 0.44(8) 0.41(2) 0.42(3) − 0.60(3) 0.44(11) 0.46(9) 0.52(7) 0.44(7) 0.42(6) 

Note: Numbers in () denote lag months, bold values indicate p-value < 0.05.

T_Max indicates average monthly maximum surface temperature, T_95 indicates 95th percentile monthly maximum surface temperature. El Niño ONI values are positive (plus values). La Niña ONI values are negative (minus values).

Table 5

Spearman's rank correlation coefficient values for IOD vs maximum surface temperature in northern and northeastern Thailand

JanFebMarAprMayJunJulAugSepOctNovDec
Positive IOD vs maximum surface temperature 
North             
T_Max 0.80(5) 0.83(4) 0.95(4)  0.80(7) 0.88(8) − 0.68(1) 0.69(4) 0.78(8) 0.78(8) 0.64(1)  
T_95 1.00(7) 0.71(6) 0.95(7)  1.00(9) 0.62(8) − 0.88(1) − 0.59(0) 0.84(8) 0.80(9) 0.69(3)  
Northeast             
T_Max 0.80(4) 0.77(4) 0.95(4)  0.80(7) 0.81(8) − 0.68(2) 0.83(4) 0.77(8) 0.82(8) 0.69(4)  
T_95 0.80(3) 0.77(6) 0.95(4)  1.00(9) 0.76(8) − 0.64(1) 0.70(4) 0.89(11) 0.73(8) 0.72(3)  
Negative IOD vs maximum surface temperature 
North             
T_Max  0.96(11) − 1.00(5) 0.77(5) 0.53(3) 0.65(4) 0.37(2) − 0.91(8) − 0.54(10) 0.32(11) 0.66(1)  
T_95  0.86(10) 0.90(8) − 0.77(11) 0.50(9) − 0.82(1) − 0.65(6) − 0.71(8) 0.38(0) 0.41(11) 0.66(5)  
Northeast             
T_Max  0.82(11) 1.00(8) 0.94(4) − 0.45(7) 0.72(9) 0.41(8) − 0.71(8) 0.52(2) 0.45(2) 0.77(1)  
T_95  0.64(11) 1.00(8) 0.71(4) − 0.38(10) − 0.60(0) 0.38(8) − 0.61(8) − 0.37(10) 0.37(5) − 0.77(9)  
JanFebMarAprMayJunJulAugSepOctNovDec
Positive IOD vs maximum surface temperature 
North             
T_Max 0.80(5) 0.83(4) 0.95(4)  0.80(7) 0.88(8) − 0.68(1) 0.69(4) 0.78(8) 0.78(8) 0.64(1)  
T_95 1.00(7) 0.71(6) 0.95(7)  1.00(9) 0.62(8) − 0.88(1) − 0.59(0) 0.84(8) 0.80(9) 0.69(3)  
Northeast             
T_Max 0.80(4) 0.77(4) 0.95(4)  0.80(7) 0.81(8) − 0.68(2) 0.83(4) 0.77(8) 0.82(8) 0.69(4)  
T_95 0.80(3) 0.77(6) 0.95(4)  1.00(9) 0.76(8) − 0.64(1) 0.70(4) 0.89(11) 0.73(8) 0.72(3)  
Negative IOD vs maximum surface temperature 
North             
T_Max  0.96(11) − 1.00(5) 0.77(5) 0.53(3) 0.65(4) 0.37(2) − 0.91(8) − 0.54(10) 0.32(11) 0.66(1)  
T_95  0.86(10) 0.90(8) − 0.77(11) 0.50(9) − 0.82(1) − 0.65(6) − 0.71(8) 0.38(0) 0.41(11) 0.66(5)  
Northeast             
T_Max  0.82(11) 1.00(8) 0.94(4) − 0.45(7) 0.72(9) 0.41(8) − 0.71(8) 0.52(2) 0.45(2) 0.77(1)  
T_95  0.64(11) 1.00(8) 0.71(4) − 0.38(10) − 0.60(0) 0.38(8) − 0.61(8) − 0.37(10) 0.37(5) − 0.77(9)  

Note: Numbers in () denote lag months, bold values indicate p-value < 0.05.

T_Max indicates average monthly temperature, T_95 indicates 95th percentile monthly temperature. Positive IOD values are positive (plus values) and negative IOD values are negative (minus values).

Table 6

Correlation coefficient values for ENSO vs minimum surface temperature in northern and northeastern Thailand

DJFJFMFMAMAMAMJMJJJJAJASASOSONONDNDJ
El Niño vs minimum surface temperature 
North             
T_Min − 0.26(11) 0.34(4) 0.37(3) 0.36(1) − 0.32(8) 0.52(10) 0.47(5) 0.56(4) 0.44(3) 0.39(2) 0.37(1) 0.34(1) 
T_05 0.44(0) 0.36(4) 0.42(11) − 0.45(8) 0.57(5) 0.60(7) 0.63(5) 0.76(4) 0.64(3) 0.56(2) 0.53(1) 0.47(1) 
Northeast             
T_Min 0.28(4) 0.31(3) − 0.44(0) − 0.41(8) − 0.57(10) − 0.27(4) 0.39(9) 0.44(4) − 0.28(1) − 0.24(11) 0.28(2) 0.32(1) 
T_05 0.52(4) 0.45(3) − 0.24(0) 0.50(1) − 0.70(10) 0.35(11) 0.64(5) 0.72(4) 0.41(3) 0.30(7) 0.46(6) 0.47(5) 
La Niña vs minimum surface temperature 
North             
T_Min 0.26(5) 0.31(4) 0.26(3) 0.41(2) − 0.41(9) 0.37(7) 0.35(6) 0.41(10) 0.31(9) 0.39(4) 0.41(2) 0.30(1) 
T_05 0.30(5) 0.27(4) 0.24(9) 0.45(9) − 0.55(0) − 0.32(4) 0.40(6) 0.49(3) 0.42(2) 0.43(1) 0.53(0) 0.41(1) 
Northeast             
T_Min 0.45(2) 0.57(1) 0.61(0) 0.35(2) 0.35(1) 0.31(7) 0.30(9) 0.37(8) 0.23(9) 0.31(3) 0.40(1) 0.47(3) 
T_05 0.42(2) 0.55(1) 0.58(0) 0.33(2) 0.40(1) − 0.32(4) 0.36(8) 0.40(3) 0.37(2) 0.43(3) 0.60(0) 0.46(1) 
DJFJFMFMAMAMAMJMJJJJAJASASOSONONDNDJ
El Niño vs minimum surface temperature 
North             
T_Min − 0.26(11) 0.34(4) 0.37(3) 0.36(1) − 0.32(8) 0.52(10) 0.47(5) 0.56(4) 0.44(3) 0.39(2) 0.37(1) 0.34(1) 
T_05 0.44(0) 0.36(4) 0.42(11) − 0.45(8) 0.57(5) 0.60(7) 0.63(5) 0.76(4) 0.64(3) 0.56(2) 0.53(1) 0.47(1) 
Northeast             
T_Min 0.28(4) 0.31(3) − 0.44(0) − 0.41(8) − 0.57(10) − 0.27(4) 0.39(9) 0.44(4) − 0.28(1) − 0.24(11) 0.28(2) 0.32(1) 
T_05 0.52(4) 0.45(3) − 0.24(0) 0.50(1) − 0.70(10) 0.35(11) 0.64(5) 0.72(4) 0.41(3) 0.30(7) 0.46(6) 0.47(5) 
La Niña vs minimum surface temperature 
North             
T_Min 0.26(5) 0.31(4) 0.26(3) 0.41(2) − 0.41(9) 0.37(7) 0.35(6) 0.41(10) 0.31(9) 0.39(4) 0.41(2) 0.30(1) 
T_05 0.30(5) 0.27(4) 0.24(9) 0.45(9) − 0.55(0) − 0.32(4) 0.40(6) 0.49(3) 0.42(2) 0.43(1) 0.53(0) 0.41(1) 
Northeast             
T_Min 0.45(2) 0.57(1) 0.61(0) 0.35(2) 0.35(1) 0.31(7) 0.30(9) 0.37(8) 0.23(9) 0.31(3) 0.40(1) 0.47(3) 
T_05 0.42(2) 0.55(1) 0.58(0) 0.33(2) 0.40(1) − 0.32(4) 0.36(8) 0.40(3) 0.37(2) 0.43(3) 0.60(0) 0.46(1) 

Note: Numbers in () denote lag months, bold values indicate p-value < 0.05.

T_Min indicates average monthly minimum surface temperature, T_05 indicates fifth percentile monthly minimum surface temperature. El Niño ONI values are positive (plus values). La Niña ONI values are negative (minus values).

Table 7

Correlation coefficient values for IOD vs minimum surface temperature in northern and northeastern Thailand

JanFebMarAprMayJunJulAugSepOctNovDec
Positive IOD vs minimum surface temperature 
North             
T_Min − 0.80(1) 0.60(3) 0.95(4)  0.80(10) 0.21(0) − 0.65(0) − 0.61(2) 0.66(8) 0.59(10) 0.59(7)  
T_05 − 0.80(1) − 0.77(7) 0.95(2)  − 0.60(11) 0.57(1) − 0.75(2) − 0.59(7) 0.65(8) 0.63(6) 0.58(7)  
Northeast             
T_Min 0.80(4) − 0.60(10) 0.74(11)  − 0.80(4) 0.90(8) − 0.52(11) 0.69(4) 0.68(8) 0.53(9) 0.90(4)  
T_05 1.00(8) − 0.71(10) 0.95(9)  − 0.60(4) 0.88(9) − 0.55(0) 0.75(8) 0.60(8) 0.51(9) 0.79(4)  
Negative IOD vs minimum surface temperature 
North             
T_Min  0.96(11) 1.00(9) 0.89(9) 0.55(0) 0.68(5) 0.57(4) − 0.42(10) 0.63(2) − 0.47(0) 0.77(2)  
T_05  0.71(11) − 0.80(4) 0.77(9) 0.69(11) 0.77(4) 0.59(8) − 0.38(10) 0.52(3) − 0.64(0) 0.77(1)  
Northeast             
T_Min  0.79(4) 0.90(8) 0.77(9) 0.46(5) 0.62(5) 0.43(4) 0.47(11) 0.58(2) − 0.62(4) 0.77(5)  
T_05  0.79(6) 0.70(0) 0.37(0) 0.52(11) − 0.72(10) 0.59(4) 0.47(11) 0.58(2) − 0.54(0) 0.83(1)  
JanFebMarAprMayJunJulAugSepOctNovDec
Positive IOD vs minimum surface temperature 
North             
T_Min − 0.80(1) 0.60(3) 0.95(4)  0.80(10) 0.21(0) − 0.65(0) − 0.61(2) 0.66(8) 0.59(10) 0.59(7)  
T_05 − 0.80(1) − 0.77(7) 0.95(2)  − 0.60(11) 0.57(1) − 0.75(2) − 0.59(7) 0.65(8) 0.63(6) 0.58(7)  
Northeast             
T_Min 0.80(4) − 0.60(10) 0.74(11)  − 0.80(4) 0.90(8) − 0.52(11) 0.69(4) 0.68(8) 0.53(9) 0.90(4)  
T_05 1.00(8) − 0.71(10) 0.95(9)  − 0.60(4) 0.88(9) − 0.55(0) 0.75(8) 0.60(8) 0.51(9) 0.79(4)  
Negative IOD vs minimum surface temperature 
North             
T_Min  0.96(11) 1.00(9) 0.89(9) 0.55(0) 0.68(5) 0.57(4) − 0.42(10) 0.63(2) − 0.47(0) 0.77(2)  
T_05  0.71(11) − 0.80(4) 0.77(9) 0.69(11) 0.77(4) 0.59(8) − 0.38(10) 0.52(3) − 0.64(0) 0.77(1)  
Northeast             
T_Min  0.79(4) 0.90(8) 0.77(9) 0.46(5) 0.62(5) 0.43(4) 0.47(11) 0.58(2) − 0.62(4) 0.77(5)  
T_05  0.79(6) 0.70(0) 0.37(0) 0.52(11) − 0.72(10) 0.59(4) 0.47(11) 0.58(2) − 0.54(0) 0.83(1)  

Note: Numbers in () denote lag months, bold values indicate p-value < 0.05.

T_Min indicates average monthly minimum surface temperature and T_05 indicates fifth Percentile monthly minimum surface temperature. Positive IOD values are positive (plus values) and negative IOD values are negative (minus values).

Figure 3

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical mixing ratio in northern Thailand.

Figure 3

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical mixing ratio in northern Thailand.

Close modal
Figure 4

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical mixing ratio in northeastern Thailand.

Figure 4

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical mixing ratio in northeastern Thailand.

Close modal
Figure 5

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical mixing ratio in northern Thailand.

Figure 5

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical mixing ratio in northern Thailand.

Close modal
Figure 6

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical mixing ratio in northeastern Thailand.

Figure 6

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical mixing ratio in northeastern Thailand.

Close modal
Figure 7

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical temperature in northern Thailand.

Figure 7

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical temperature in northern Thailand.

Close modal
Figure 8

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical temperature in northeastern Thailand.

Figure 8

Correlation coefficient (ρ) heat maps of El Niño vs 300, 500, 700, 850 hPa and vertical temperature in northeastern Thailand.

Close modal
Figure 9

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical temperature in northern Thailand.

Figure 9

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical temperature in northern Thailand.

Close modal
Figure 10

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical temperature in northeastern Thailand.

Figure 10

Correlation coefficient (ρ) heat maps of La Niña vs 300, 500, 700, 850 hPa and vertical temperature in northeastern Thailand.

Close modal

Teleconnections vs rainfall

ENSO vs rainfall

The correlations between ENSO and monthly rainfall (R_M) and 95th percentile rainfall (R_95) in northern and northeastern Thailand were found moderately correlated. However, El Niño found more associations in northern than in northeast Thailand. In northern Thailand, El Niño occurred at JFM season was correlated at −0.54 with April R_M, which is similar to the findings of Ueangsawat (2013) of JFM SOI indices at 0.51 with May rainfall. In the northeast, El Niño occurred during the MAM season, and March R_M correlated at −0.75, and El Niño AMJ correlated with September R_95 at −0.57, indicating that El Niño may be associated with R_M and R_95 rainfall, during the dry season and the dry spell. However, northern and northeastern Thailand also found a positive correlation in some months.

There were not so many correlations between La Niña with R_M and R_95 in northern and northeastern Thailand. La Niña state is represented with negative values, therefore the negative direction of correlation between La Niña and rainfall means that when La Niña is stronger, rainfall can be higher than usual. In this study, only moderate correlations (−0.60) were found in La Niña FMA and October R_M, and La Niña MJJ and August R_M in northern Thailand as depicted in Table 2.

The above results of ENSO vs rainfall are similar to those of past studies in other Asian areas. Räsänen & Kummu (2013) found a moderate negative correlation between ENSO and rainfall in the Mekong basin, especially in northeast Thailand. Mishra et al. (2020) also found a moderate negative correlation between observed rainfall and ENSO variability in India. Moreover, the monthly lag times between La Niña events and rainfall in Korea and Japan were found at 4–5 lag months (Kawamura et al. 2005). In addition, extreme rainfall which caused devastating floods in Australia during 2010–2011 was found to have a strong link with La Niña events (Ummenhofer et al. 2015).

IOD vs rainfall

The correlation between positive IOD and rainfall in northern and northeastern Thailand is found to be negatively moderate, similar to El Niño. as depicted in Table 3. In northern Thailand, July–November p_IOD shows a moderate correlation with next year's R_M and R_95 in the dry season (February and April), early rainy season (May), rainy season (September), and late rainy season (October). Similarly, in the northeast, July–October p-IOD is associated with February, April, November, R_M, and February, April R_95, which may be associated with droughts in these areas. Cold phases or negative IOD (n-IOD) and R_M, R_95 also show moderate positive and negative associations. In contrast, Bridhikitti (2019) found no significant correlation between IOD and northern rainfall in May and April, however, IOD likely co-occurred with ENSO. In addition, the results of Ummenhofer et al. (2013) indicated that the co-occurrence between positive IOD and El Niño events could have worsened severe droughts in the Indo-Pacific region from 1877 to 2006.

MJO vs rainfall

Negligible values of correlation coefficient (ρ less than 0.2) were found between MJO and weekly rainfall. These results differ from the moderate −0.54 correlation coefficient values between MJO and May rainfall in northern Thailand (Bridhikitti 2019). Furthermore, Xavier et al. (2014) found a relationship between increased (decreased) extreme rainfall probability and convective (suppressive) MJO phases in Southeast Asia. Details of the correlation coefficient for MJO vs rainfall are depicted in the Supplementary Material, Table S7.

Teleconnections vs maximum surface temperature

The results of the correlation between teleconnection indices and monthly maximum surface temperature (T_Max) and 95th percentile monthly maximum surface temperature (T_95) are depicted in Tables 4 and 5 in the following sections.

ENSO vs maximum surface temperature

The positive correlations between El Niño, T_Max, and T_95 in northern and northeastern Thailand indicate greater warmth than usual in the dry season (April, March) and the rainy season (May and August) as depicted in Table 4. The increasing maximum surface temperature can intensify the evaporation process, whereas La Niña can be associated with daytime temperature reduction (Bureau of Meteorology 2021b). The positive direction of La Niña vs T_Max, T_95 shows that when La Niña is stronger, the T_Max and T_95 will be colder (see the details of correlation directions in the Supplementary Material, Table S3). A moderate positive correlation (0.4–0.6) was found between La Niña in April and May T_Max in northern Thailand. In northeastern Thailand, positive correlations were mostly found between La Niña and the dry season (February, April) and the rainy season (June, October) of T_Max.

IOD vs maximum surface temperature

The positive IOD (p_IOD) can influence higher-than-usual temperature, in contrast with negative IOD (n_IOD). The p_IOD are more correlated to T_Max and T_95 than n_IOD, as depicted in Table 5. In northern Thailand, the dry season (December), the rainy season (May), and the dry spell month (June) T_Max were positively correlated with p_IOD, whereas a negative correlation can be found between p_IOD and the rainy season (August) T_Max and T-95. The positive correlation in northeastern Thailand was also found between p_IOD and December, May, June T_Max. However, the n_IOD has both negative and positive correlations with T_Max and T_95 in northern and northeastern Thailand. The moderate to high negative correlation values were found between August n_IOD and April T_Max and T_95 in both northern and northeastern Thailand.

MJO vs maximum surface temperature

MJO and weekly maximum surface temperature also show a mostly positive correlation in both phases. However, the highest magnitude of the correlation is less than 0.2, which is negligible. Details of the correlation coefficient for MJO vs maximum surface temperature are depicted in the Supplementary Material, Table S8.

Teleconnections vs minimum surface temperature

The correlation coefficient results between teleconnection indices and minimum surface temperature (T_Min) and fifth percentile minimum surface temperature (T_05) are depicted in Tables 6 and 7.

ENSO vs minimum surface temperature

El Niño and La Niña have both moderate positive and negative correlations with T_Min and T_05 in northern and northeastern Thailand as depicted in Table 6. The positive correlation between El Niño and T_Min implies that when El Niño is stronger, T_Min is warmer. In contrast, a positive correlation between La Niña and T_Min implies that when La Niña is stronger, T_Min is colder. The positive moderate correlation between El Niño and T_Min, T_05 was found in northern Thailand more than in northeastern Thailand. In northern Thailand, a moderate positive correlation was found between El Niño and March, November T_Min. For T_05, a positive correlation was found between El Niño and September, November, and December T_05. For northeastern Thailand, a positive correlation was found between April and November T_05. The correlation between La Niña and October, December T_Min and T_05 in northern Thailand was also positive. In northeastern Thailand, a positive correlation was found between La Niña and February T_Min, and February, October, and December T_05.

IOD vs minimum surface temperature

The positive IOD (p_IOD) and negative IOD (n_IOD) correlate positively and negatively with T_Min and T_05 in northern and northeastern Thailand, similar to ENSO. A moderate positive correlation in northern Thailand was found between p_IOD and May, June, and August T_Min, and between p_IOD and April, May, and June T_05. In northeastern Thailand, the p_IOD was found moderately correlated with March, May, December T_Min, and March, April T_05. The n_IOD was found moderately correlated with May, November T_Min, and March, December T_05 in northern Thailand. In northeastern Thailand, the n_IOD was found positively correlated with November T_Min, T_05.

MJO vs minimum surface temperature

MJO and T_Min also show correlation coefficient values less than 0.2 but with mostly negative correlation. Details of correlation coefficient results for MJO vs minimum surface temperature are depicted in the Supplementary Material, Table S9.

ENSO vs vertical humidity

Vertical humidity is one of the crucial components of cloud formation. The surface to 700 hPa level is the level at which clouds form and grow, therefore the variation of vertical humidity at this level can affect cloud amounts and might affect rainfall variation. The correlation coefficient (CC) values between El Niño, monthly mixing ratio (MR), and 95th percentile monthly mixing ratio (MR_95) in northern Thailand are shown in Figure 3. Moderate negative correlations were found in March and April MR, and it coincides with previous results of negative correlation between El Niño and April R_M and El Niño and April T_Max positive correlations. The highest CC value −0.71 was found at the surface level of the 95th percentile monthly mixing ratio (MR_95) in March.

In northeastern Thailand, the correlation between El Niño and MR was mainly positive as depicted in Figure 4. The moderate negative correlation between El Niño and MR_95 was rarely found in northeastern Thailand, which contrasts with northern Thailand. The positive correlation between El Niño and MR implies that when El Niño is strong, MR also increases. However, not only MR amounts but also other variables should be considered for clouds and rainfall variation, such as temperature and wind patterns. Furthermore, mixing ratio data was calculated from only one point at the upper-air station, therefore the area and distance from the ocean may also be affected by the correlations.

The correlation results between La Niña and MR, MR_95 in northern Thailand, showed a more negative correlation than El Niño as depicted in Figure 5. At 850 hPa level, moderate negative correlations were found with March, April MR, and April MR_95. Trakolkul et al. (2022) indicated that La Niña events can induce water vapor increase in Thailand, especially as during the severe floods of 2011. However, the El Niño event has no relationship with water vapor, contrary to our results.

The correlation results between La Niña, MR, and MR_95 in northeastern Thailand are depicted in Figure 6. A negative correlation was found between La Niña and MR; MR_95 is negatively correlated between AMJ La Niña and April, June, October MR. In addition, a negative correlation between AMJ La Niña and April, October MR_95 was also presented.

ENSO vs vertical temperature

The correlation results between ENSO, monthly vertical temperature (VT), and 95th percentile monthly vertical temperature (VT_95) were found more correlated with La Niña than El Niño in northern and northeastern Thailand. The vertical temperatures at 300 and 500 hPa are lower than 0 °C, thus their values are negative as described in the Supplementary Material, Table S3. The results between El Niño, VT, and VT_95 in northern Thailand are depicted in Figure 7.

The negative correlation between El Niño and August VT at 300 and 500 hPa levels indicates an inverse relationship, which can imply that stronger El Niño can be associated with colder August VT. In contrast, a moderate positive correlation is also found between El Niño and April VT, which can imply that stronger El Niño can be associated with warmer April VT in the higher level.

A positive correlation between El Niño and Sfc April VT and February VT is presented in northeastern Thailand. In addition, El Niño and 700 hPa March VT also show a positive correlation as depicted in Figure 8. The positive correlation implies that when El Niño is stronger, VT is colder in the dry season. A negative correlation is also presented between El Niño and June VT_95 at the 850 hPa level, which implies colder June VT_95 when El Niño is strong.

La Niña events can be associated with the decreasing of vertical temperature. At the 300 and 500 hPa levels, the positive correlation direction indicates decreasing temperature when La Niña is stronger. Moreover, the significant results between La Niña, VT, and VT_95 were found to correlate during the rainy season in August, May, June, and September in northern Thailand as shown in Figure 9.

The positive correlation results between La Niña and VT, VT_95 in northeastern Thailand were found more significant at higher levels, which implies the inverse relationship between La Niña and higher-level VT in June. In contrast to northern Thailand, the significant correlations of northeastern Thailand were found mostly during June, December, and October as depicted in Figure 10.

Spatiotemporal influence of ENSO on average rainfall anomalies

The spatial distributions of rainfall in both the onset and decay years of ENSO events are presented in Figures 11 and 12. In addition, the ENSO intensities between 1951 and 2021 were classified into three groups: strong, moderate, and weak, as per the Supplementary Material, Table S4. The results clearly show that the strong El Niño had more influence on rainfall anomalies in the post-1980 period than pre-1980 in both onset and decay years. In contrast, the moderate El Niño maps show insignificant differences between pre- and post-1980 anomaly rainfall in onset years. However, in the decay years, the post-1980 anomaly rainfall percentage is higher than that pre-1980. Moreover, in the onset years of weak El Niño, pre-1980 had more negative rainfall anomalies than in post-1980. However in decay years, the post-1980 negative rainfall anomalies are higher than those that pre-1980. The details of rainfall anomaly maps during El Niño events are depicted in Figure 11.
Figure 11

Average annual rainfall anomalies (%) of the (a) strong El Niño, (b) moderate El Niño, (c) weak El Niño years during 1951–1980 and 1981–2020. The number (0) denotes the onset year of El Niño events and (1) denotes the decay year of El Niño events.

Figure 11

Average annual rainfall anomalies (%) of the (a) strong El Niño, (b) moderate El Niño, (c) weak El Niño years during 1951–1980 and 1981–2020. The number (0) denotes the onset year of El Niño events and (1) denotes the decay year of El Niño events.

Close modal
Figure 12

Average annual rainfall anomalies (%) of the (a) strong La Niña, (b) moderate La Niña, (c) weak La Niña years during 1951–1980 and 1981–2020. The number (0) denotes the onset years of La Niña events and (1) denotes the decay years of La Niña events.

Figure 12

Average annual rainfall anomalies (%) of the (a) strong La Niña, (b) moderate La Niña, (c) weak La Niña years during 1951–1980 and 1981–2020. The number (0) denotes the onset years of La Niña events and (1) denotes the decay years of La Niña events.

Close modal

Average annual rainfall anomaly maps depicted in Figure 12 are based on La Niña intensities in onset and decay years during pre- and post-1980. Despite this, the strong La Niña show more positive rainfall anomalies in the pre-1980 period than those of post-1980. The decay years of the strong La Niña show more positive rainfall anomalies in post-1980 in both northern and northeastern Thailand. In contrast, moderate La Niña shows higher positive rainfall anomalies in onset years and weak La Niña shows higher positive rainfall anomalies in both onset and decay years in post-1980. However, the decay years of moderate La Niña show higher positive rainfall anomalies in pre-1980.

The two periods of (1) 1951–1980 and (2) 1981–2021 were used to investigate the frequency and intensity of ENSO events as per the details in the Supplementary Material, Table S4. During both periods, there were a total of 26 El Niño events, period (1) 13 events which are equal to period (2). The strong El Niño in period (1) was three events and in period (2) was five. Moreover, in period (1) the moderate El Niño occurred three times, and four times for period (2). Therefore, the intensities of El Niño events were slightly increasing during period (2). In contrast, the frequency and intensity of La Niña events during period (2) were increasing. The La Niña events during period (1) were eight, whereas the La Niña events during period (2) were doubled to 16. Furthermore, the strong La Niña were two and five during periods (1) and (2), respectively. In addition, moderate La Niña events increased from two events in period (1) to four events in period (2).

ENSO impact on different elevations

To investigate the implication of ENSO events for different topographies, the 21 TMD weather stations were classified into two groups: mountainous and plain area groups. The contour maps of rainfall anomalies during 1951–2021 based on El Niño events in Figure 13(a) and La Niña events in Figure 13(b) show trivial differences between the two groups. The percentage of rainfall anomalies during El Niño events in the mountainous area is between 2% and −2%, and in the plain area it varies between 3% and −9%. Furthermore, during La Niña events, the percentage of rainfall anomalies in the mountainous area is between 8% and 0%, and in the plain area it varies between 8% and 2%.
Figure 13

Average annual rainfall anomaly contour maps (%) of the (a) El Niño events and (b) La Niña events during 1951–2020.

Figure 13

Average annual rainfall anomaly contour maps (%) of the (a) El Niño events and (b) La Niña events during 1951–2020.

Close modal

To establish the relationship between major teleconnections and regional weather of northern and northeastern Thailand, the lag correlations between ENSO, IOD, MJO indices, and 21 TMD weather stations' rainfall, maximum, and minimum surface temperature data and two upper-air stations' vertical humidity and average temperature data were performed. Subsequently, the statistical significance was performed using a t-test at a 95% confidence level (p-value < 0.05). The results clearly showed that ENSO and IOD have a moderate relationship with climate variables. The El Niño or warm–dry phase of ENSO showed a negative significant relationship with monthly rainfall at the highest values of −0.73 and −0.75 in northern and northeastern Thailand, respectively. Furthermore, El Niño showed a significant negative relationship with 95th percentile monthly rainfall at the highest values of −0.71 and −0.75 in northern and northeastern Thailand, respectively. Moreover, La Niña depicted a more significant relationship with rainfall in northern than in northeastern Thailand. The highest value of the correlation coefficient is −0.60.

The relationship between positive IOD and rainfall showed the highest value of −0.75 for northern and −0.71 for northeastern Thailand. In contrast, MJO's significant correlation values were negligible. The significant correlations between El Niño, positive IOD and rainfall, maximum temperature, vertical mixing ratios, and vertical temperature demonstrated the probability of severe drought exacerbation in a dry spell and dry season. The moderately significant correlations between El Niño and vertical mixing ratios with the highest values of −0.63 in March indicate that they coincide with El Niño and rainfall showing significant correlations in March and April in northern Thailand.

The significant correlations between La Niña, negative IOD, and regional weather variabilities were found less than in El Niño, positive IOD, and found more in northern than in northeastern Thailand. The results suggest a potential for improving teleconnection-induced meteorological event forecasts in northern and northeastern Thailand. Since teleconnections may exacerbate extreme events such as droughts, additional policy planning for water resource management, extreme events risk management, and agriculture should be considered to reduce adverse consequences on the environment, economics, and livelihoods.

Future research could further examine the relationship between teleconnections and climate variables using additional climate models or ensemble data, especially vertical water vapor data for more reliable patterns. Further advanced statistical methods should be applied to find more associations and impacts between teleconnections and MJO vs regional weather such that robust forecast modeling with ensemble teleconnections can be established.

The Ministry of Agricultural and Cooperatives, Thailand, supports the first author through a graduate scholarship. Furthermore, the first author would like to thank Mr Chanti Detyothin, Expert on Applied Atmospheric Science Research and Development, Department of Royal Rainmaking and Agricultural Aviation, Ministry of Agricultural and Cooperatives, Thailand, who provided valuable insights and expertise that greatly assisted the research. The authors would like to thank Thai Meteorological Department and Hydrology Division, Office of Water Management and Hydrology, Royal Irrigation Department Thailand for the valuable climate data.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Aamir
E.
&
Hassan
I.
2020
The impact of climate indices on precipitation variability in Baluchistan, Pakistan
.
Tellus A: Dynamic Meteorology and Oceanography
72
(
1
),
1833584
.
Aamir
E.
,
Khan
A.
&
Abubakar Tariq
M.
2022
The influence of teleconnections on the precipitation in Baluchistan
.
Atmosphere
13
(
7
),
1001
.
AMS
2022
mixing ratio. Available from: https://glossary.ametsoc.org/wiki/Mixing_ratio (accessed 12 July)
.
Boers
N.
,
Goswami
B.
,
Rheinwalt
A.
,
Bookhagen
B.
,
Hoskins
B.
&
Kurths
J.
2019
Complex networks reveal global pattern of extreme-rainfall teleconnections
.
Nature
566
(
7744
),
373
377
.
Bridhikitti
A.
2019
Multi-decadal trends and oscillations of Southeast Asian monsoon rainfall in northern Thailand
.
Songklanakarin Journal of Science & Technology
41
(
1
),
74
80
.
Bureau of Meteorology
2021a
Indian Ocean influences on Australian climate. Bureau of Meteorology, Australian Government. http://www.bom.gov.au/climate/iod/ (accessed 21 March 2021)
.
Bureau of Meteorology
2021b
ENSO impacts – temperature. Bureau of Meteorology, Australian Government. http://bom.gov.au/climate/enso/history/ln-2010-12/ENSO-temperature.shtml (accessed 21 March 2021)
.
Cash
B. A.
,
Barimalala
R.
,
Kinter
J. L.
,
Altshuler
E. L.
,
Fennessy
M. J.
,
Manganello
J. V.
,
Molteni
F.
,
Towers
P.
&
Vitart
F.
2017
Sampling variability and the changing ENSO–monsoon relationship
.
Climate Dynamics
48
(
11–12
),
4071
4079
.
Ceglar
A.
,
Turco
M.
,
Toreti
A.
&
Doblas-Reyes
F. J.
2017
Linking crop yield anomalies to large-scale atmospheric circulation in Europe
.
Agricultural and Forest Meteorology
240–241
,
35
45
.
Ciemer
C.
,
Rehm
L.
,
Kurths
J.
,
Donner
R. V.
,
Winkelmann
R.
&
Boers
N.
2020
An early-warning indicator for Amazon droughts exclusively based on tropical Atlantic sea surface temperatures
.
Environmental Research Letters
15
(
9
),
094087
.
Clarke
B.
,
Otto
F.
,
Stuart-Smith
R.
&
Harrington
L.
2022
Extreme weather impacts of climate change: an attribution perspective
.
Environmental Research: Climate
1
(
1
),
012001
.
Cortés
J.
,
Mahecha
M.
,
Reichstein
M.
&
Brenning
A.
2020
Accounting for multiple testing in the analysis of spatio-temporal environmental data
.
Environmental and Ecological Statistics
27
(
2
),
293
318
.
Czymzik
M.
,
Muscheler
R.
&
Brauer
A.
2016
Solar modulation of flood frequency in Central Europe during spring and summer on inter-annual to millennial time-scales
.
Climate of the Past Discussions
11
,
4833
4850
.
Feldstein
S. B.
&
Franzke
C. L. E.
2017
Atmospheric teleconnection patterns
. In:
Nonlinear and Stochastic Climate Dynamics
.
(Franzke, C. L. E. & O'Kane, T. J., eds)
,
Cambridge University Press
,
Cambridge, UK
, pp.
54
104
.
https://doi.org/10.1017/9781316339251.004
.
Golden Gate Weather Services
2022
El Niño and La Niña Years and Intensities
.
Available from: https://ggweather.com/enso/oni.htm (accessed 9 February 2023)
.
Gonsamu
A.
&
Chen
J. M.
2015
Winter teleconnections can predict the ensuing summer European crop productivity
.
Proceedings of the National Academy of Sciences of the United States of America
112
(
18
),
E2265
E2266
.
IPCC
2021
Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (Masson-Delmotte, V., Zhai, P., Pirani, A., Connors, S. L., Péan, C., Berger, S., Caud, N., Chen, Y., Goldfarb, L., Gomis, M. I., Huang, M., Leitzell, K., Lonnoy, E., Matthews, J. B. R., Maycock, T. K., Waterfield, T., Yelekçi, O., Yu, R. & Zhou, B., eds). Cambridge University Press, Cambridge, UK and New York, USA
.
Irwandi
H.
,
Pusparini
N.
,
Ariantono
J. Y.
,
Kurniawan
R.
,
Tari
C. A.
&
Sudrajat
A.
2018
The influence of ENSO to the rainfall variability in North Sumatra Province
.
IOP Conference Series: Materials Science and Engineering
335
,
012055
.
Kawamura
A.
,
Jin
Y. H.
,
Jinno
K.
&
Berndtsson
R.
2005
Impacts of ENSO on monthly precipitation in South Korea and Fukuoka, Japan. In: Regional Hydrological Impacts of Climatic Change – Hydroclimatic Variability (Franks, S., Wagener, T., Bøgh, E., Gupta, H. V., Bastidas, L., Nobre, C. & De Oliveira Galvão, C., eds), IAHS Publication 296, IAHS Press, Wallingford, UK, pp. 246–254
.
Lee
J. H.
,
Julien
P. Y.
&
Lee
S.
2023
Teleconnection of ENSO extreme events and precipitation variability over the United States
.
Journal of Hydrology
619
,
129206
.
Lim
H. S.
&
Boochabun
K.
2012
Flood generation during the SW monsoon season in northern Thailand
.
Geological Society, London, Special Publications
361
,
7
20
.
Liu
T.
,
Chen
D.
,
Yang
L.
,
Meng
J.
,
Wang
Z.
,
Ludescher
J.
,
Fan
J.
,
Yang
S.
,
Chen
D.
,
Kurths
J.
,
Chen
X.
,
Havlin
S.
&
Schellnhuber
H. J.
2023
Teleconnections among tipping elements in the Earth system
.
Nature Climate Change
13
,
67
74
.
Lüdecke
H.-J.
,
Müller-Plath
G.
,
Wallace
M. G.
&
Lüning
S.
2021
Decadal and multidecadal natural variability of African rainfall
.
Journal of Hydrology: Regional Studies
34
,
100795
.
MetOffice
2011
Fact Sheet No. 13 – Upper Air Observations and the Tephigram
. .
NOAA
2020
El Niño Southern Oscillation (ENSO)
.
Physical Sciences Laboratory, NOAA. Available from: https://psl.noaa.gov/enso/ (accessed 12 July 2022)
.
Nounmusig
W.
2018
Analysis of rainfall in the eastern Thailand
.
International Journal of GEOMATE
14
(
46
),
150
155
.
NSO
2013
Agricultural Census
.
National Statistical Office, Bangkok, Thailand. Available from: http://www.nso.go.th/sites/2014en/Pages/Census/Agricultural%20Census/2013/whole%20kingdom/Full%20Report.pdf (accessed 3 March 2023)
.
Phillips
T.
,
Nerem
R. S.
,
Fox-Kemper
B.
,
Famiglietti
J. S.
&
Rajagopalan
B.
2012
The influence of ENSO on global terrestrial water storage using GRACE
.
Geophysical Research Letters
39
(
16
),
L16705
.
https://doi.org/10.1029/2012GL052495.
Promchote
P.
,
Wang
S.-Y. S.
&
Johnson
P. G.
2016
The 2011 great flood in Thailand: climate diagnostics and implications from climate change
.
Journal of Climate
29
(
1
),
367
379
.
Puttrawutichai
S.
,
Kwanyuen
B.
,
Kongjun
T.
,
Ishii
A.
&
Satoh
M.
2017
Relationship between El Niño-Southern oscillation and seasonal rainfall in Thailand's northern and central regions
.
Journal of Science and Technology Kasetsart University
6
(
1
),
48
61
.
Rodríguez-Fonseca
B.
,
Suárez-Moreno
R.
,
Ayarzagüena
B.
,
López-Parages
J.
,
Gómara
I.
,
Villamayor
J.
,
Mohino
E.
,
Losada
T.
&
Castaño-Tierno
A.
2016
A review of ENSO influence on the North Atlantic. A non-stationary signal
.
Atmosphere
7
(
7
),
87
.
Sahu
N.
,
Yamashiki
Y.
&
Takara
K.
2010
Impact assessment of IOD/ENSO in the Asian region
.
京都大学防災研究所年報. (Annuals of the Disaster Prevention Research Institute, Kyoto University)
53
(
B
),
97
103
.
Sattar
M. A.
,
Mia
S.
,
Shanta
A. A.
,
Biswas
A. K. M. A. A.
&
Ludwig
F.
2021
Remote impacts from El Niño and La Niña on climate variables and major crops production in coastal Bangladesh
.
Atmosphere
12
(
11
),
1449
.
Singhrattna
N.
,
Rajagopalan
B.
,
Clark
M.
&
Krishna Kumar
K.
2005
Seasonal forecasting of Thailand summer monsoon rainfall
.
International Journal of Climatology
25
(
5
),
649
664
.
Sreekala
P. P.
,
Rao
S. V. B.
,
Rajeevan
K.
&
Arunachalam
M. S.
2018
Combined effect of MJO, ENSO and IOD on the intraseasonal variability of northeast monsoon rainfall over south peninsular India
.
Climate Dynamics
51
(
9–10
),
3865
3882
.
Thai Meteorological Department
2015
The Climate of Thailand
.
Thai Meteorological Department, Thailand. Available from: https://www.tmd.go.th/en/archive/thailand_climate.pdf (accessed 2 November 2020)
.
Trakolkul
C.
,
Charoenphon
C.
&
Satirapod
C.
2022
Impact of El Niño–Southern Oscillation (ENSO) on the precipitable water vapor in Thailand from long term GPS observation
.
International Journal of Geoinformatics
18
(
3
),
13
20
.
https://doi.org/10.52939/ijg.v18i3.2197
.
Ueangsawat
K.
2013
The impacts of ENSO phases on the variation of rainfall and stream flow in the Upper Ping River Basin, northern Thailand
.
Environment and Natural Resources Journal
11
(
2
),
97
119
.
Ummenhofer
C. C.
,
D'Arrigo
R. D.
,
Anchukaitis
K. J.
,
Buckley
B. M.
&
Cook
E. R.
2013
Links between Indo-Pacific climate variability and drought in the Monsoon Asia Drought Atlas
.
Climate Dynamics
40
(
5–6
),
1319
1334
.
Ummenhofer
C. C.
,
Sen Gupta
A.
,
England
M. H.
,
Taschetto
A. S.
,
Briggs
P. R.
&
Raupach
M. R.
2015
How did ocean warming affect Australian rainfall extremes during the 2010/2011 La Niña event?
Geophysical Research Letters
42
(
22
),
9942
9951
.
UNDP
2023
Thailand
.
Climate Change Adaptation, United Nations Development Programme. Available from: https://www.adaptation-undp.org/explore/south-eastern-asia/thailand (accessed 5 February 2023)
.
UNICEF
2023
UNICEF report finds children in Thailand at ‘high risk’ from climate change and environmental degradation. UNICEF press release (28 March)
.
Wikarmpapraharn
C.
&
Kositsakulchai
E.
2010
Relationship between ENSO and rainfall in the Central Plain of Thailand
.
Agriculture and Natural Resources
44
(
4
),
744
755
.
Williams
N. C.
,
Scaife
A. A.
&
Screen
J. A.
2023
Underpredicted ENSO teleconnections in seasonal forecasts
.
Geophysical Research Letters
50
(
5
),
e2022GL101689
.
World Bank Group
2021
Thailand
.
Climate Change Knowledge Portal, World Bank Group. Available from: https://climateknowledgeportal.worldbank.org/country/thailand/vulnerability (accessed 5 February 2023)
.
Xavier
P.
,
Rahmat
R.
,
Cheong
W. K.
&
Wallace
E.
2014
Influence of Madden-Julian Oscillation on Southeast Asia rainfall extremes: observations and predictability
.
Geophysical Research Letters
41
(
12
),
4406
4412
.
Zaiontz
C.
2022
Spearman's Rank Correlation
.
Real Statistics Using Excel. Available from: https://www.real-statistics.com/correlation/ spearmans-rank-correlation/ (accessed 21 July 2021)
.
Zavadoff
B.
&
Arcodia
M.
2022
What are teleconnections? Connecting Earth's climate patterns via global information superhighways. Climate.gov (22 December). Available from: https://www.climate.gov/news-features/blogs/enso/what-are-teleconnections-connecting-earths-climate-patterns-global (accessed 9 February 2023)
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).

Supplementary data