Statistical and Spatial Analysis of Extreme Temperatures in Libya (1961-2020)

The present paper seeks to examine spatial and temporal variability of changes in both warm and cold extreme temperatures. Additionally, it aims to determine the most extreme geographical areas in extreme temperatures based on climate data recorded during the period from 1961 to 2020. This paper analysed the annual data of extreme temperatures for 16 meteorological stations during 1961-2020 across Libya over two climatic regions (coastal and inland) based on Mann–Kendall test. Time trend of the extreme of both maximum and minimum temperature will be analyses and the results from geographical analysis are will also be presented based on ArcMap. The results indicated that a significant increase in the coldest extreme (TNn) temperature are found in autumn and spring and winter at majority of stations. Temperature variability can be associated with variations in large-scale atmospheric patterns represented by Eastern Atlantic and the Western Mediterranean Oscillations.


‫لتطرفات‬ ‫والمكاني‬ ‫اإلحصائي‬ ‫التحليل‬ ( ‫ليبيا‬ ‫في‬ ‫الحرارة‬ ‫درجات‬ 1691 -0202 )
Introduction Climate extremes can be analysed by estimating changes in extreme weather and climate parameters, which are divided by Alexander et al. [1] into four different index classes: percentile-based, absolute, duration and threshold-based, with regional and global (e.g., [2,3]).Percentile-based indices represent the amount (percentage) of temperature falling above a fixed value.Absolute indices refer to counts of days crossing a specified absolute value that can be a minimum or maximum value within a month, season or year.Threshold-based indices consist of a number of days on which temperature/precipitation values fall above or below a fixed threshold.Duration indices define periods of extreme warmth, cold, wetness, or dryness.From the internationally agreed WMO list of approximately 50 climate change indices available online at htt://www.knmi.nl/samen/ecaand recommended by Jones et al. [4][5][6].Studying changes in extreme temperature, as well as attributing and predicting those changes, has great importance.For example, extreme temperature events can seriously affect human health, ecosystems, and the economy [7,8].The heat waves that swept northwestern Libya in 29 the past two years (2020, 2021), which lasted for more than ten continuous days, caused the production of many crops to deteriorate, as well as drought and the death of many trees (Figure 1).Given their impacts, it is important to understand the mechanisms that trigger those cold extreme events and improve their forecasts.The most important climate indices related to temperature and those most widely used for the analysis of extremes in temperature will be analysed in this study.

Literature Review
The Fifth Assessment Report of the IPCC [9] indicated that the surface temperature showed a change tendency rate of increase from 1880 to 2012, with an increase of 0.85 °C [9].Studies of observational temperature records over the last 50-100 years have found evidence for an increase in both observed mean, minimum and maximum temperature [10,5].In the global scope, the highest trend of the lowest temperature indices, the increase in the extreme temperature indices and the decrease in the cold extreme value all show obvious warming trends from the 20th century to the beginning of the 21st century [1,10].Morak et al. [11] indicated an increase in warm extremes and a decrease in cold extremes in both seasons and in almost all regions that are generally well captured by the model.Ageena indicated that a mix of increasing and decreasing trends on the warmest day are also found.Zhou et al. [13] showed that the extreme temperature warming indices and the minimum daily maximum temperature (TXn) and the minimum daily minimum temperature (TNn) of cold indices showed an increasing trend from 1960 to 2016, especially since the 1990s.Rodríguez et al. [14] found a significant trend of more warm days and fewer cold nights.Moreover, an increase in the number of heat waves has also been observed in southwestern North America (Tebaldi et al. [15].Shen et al. [16] found an upwards tendency in the occurrence of warm days (TX90) and nights (TN90) and in the temperatures of the coldest day (TXn), coldest night (TNn), warmest day (TXx), and warmest night (TNx) in China and most climate regions.In most extreme temperature indices in Asia, the number of warm days and nights is increasing, and the increasing speed of the latter is significantly higher than that of the former [ 7 1 ].In the past 60 years in Central Asia and North America, the growth rate of the highest temperature (TMAX) was faster than that of the lowest temperature (TMIN) [11], [19].The rise rate of the lowest temperature in Iran is twice that of the highest temperature.Therefore, the frequency of extremely high temperatures increases and the frequency of extremely low temperatures decreases [20, [10, 14].Toreti and Desiato [21] confirmed that most of the indices, averaged over all stations, showed a cooling trend until the end of the 1970s followed by a more pronounced warming trend in the last 25 years.Collazo et al. [22] found persistence in the winter cold night series compared with the previous season in eastern Argentina.In contrast to the observed increase in warm extremes, Portmann et al. [23] and Meehl et al. [24] found that the change in the number of hot daytime extremes across eastern North America was modest, with decreases over the southeaster part.

Methodology
The present paper seeks to examine the spatial and temporal variability of changes in both warm and cold extreme temperatures.Specifically, it aims to identify temporal fluctuations and patterns in extreme temperatures across Libya over coastal and inland climatic regions based on annual data.The study will be based on maximum and minimum temperature data for 16 synoptic meteorological stations for the time period 1961-2020 distributed over coastal and interior regions in Libya.All the data were obtained from the Libyan National Meteorological Centre.The study will be based on the Mann-Kendall test for warm extreme temperatures (warm days (TX90), warm nights (TN90), warmest day (TXx) and warmest night (TNx)).In addition to cold extreme temperature, cold days (TX10), cold nights (TN10), coldest days (TXn) and coldest nights (TNn) were based on indices of the CLIVAR set [21] (Table 1).The statistical significance levels used are *** = 0.01 (confidence level of 99%), ** = 0.05 (confidence level of 95%), and * = 0.10 (confidence level of 90%).Note: all indices are calculated based on TX as the daily maximum temperature and TN as the daily minimum temperature In addition, the inverse distance weighted (IDW) tool in ArcMap was used as an interpolation tool to predict the extreme trend values of cells at locations that lack sampled points.It involves using known z values and weights determined as a function of distances between the unknown and known points.As such, IDW points that are far away have far less influence than points that are close [25].

Annual variations and trends of warm extreme temperatures
The temporal and spatial patterns of changes and trends for warm extremes for warmest day (TXx), warm days (TN90p), warmest night (TNx) and warm nights (TX90p) are analysed.Regional averages of (TXx) showed positive trends at the majority of stations with an average of 0.24 °C/year with significant positive trends recorded at 10 of 16 stations with different confidence rates (Table 2).The remaining eastern coastal areas, Binina, Shahat and Darnah: Binina, indicated no trends.The increases in TXx are significant (**) at all western and central coastal stations, Zwarah, Nalut and Musratah, and at inland stations, Al-Jaghbub, Sabha and Tazerbou (Figure 2).The results show that low positive trends were recorded for the warmest night (TNx) at all stations, which ranged from 0.001 day/year (Binina) to 0.029 day/year (Zwarah), with a significant positive trend at 10 out of 16 stations (Figure 4).High significant negative trends with an average of (-0.026 day/year) in warm nights (TN90p) are identified at Zwarah, Binina, Shahat, AlJaghbub and Al-Kufrah, which ranged from -0.040 day/year to -0.138 day/year (Table 2 and Figure 3).In contrast, there is no change at 15 out of 16 stations in TX90p, except for Al-Jaghbub, which has negative changes of -0.043 day/year.

Annual variation and trends of cold extreme temperatures
This section presents the estimation of changes and trends in extreme cold temperatures for the 16 stations (1961-2020).Trends in the coldest night (TNn), coldest day (TXn), cold night percentile (TN10p) and cold day percentile (TX10p) index anomalies for 1961-2020 are analysed in this paper.Positive trends in the coldest night (TNn) are found at all stations, ranging from 0.006 °C/year at Tazerbou to 0.091 °C/year at Ghadames (Figure 5), with significant increases at all stations except Tazerbou, with a highly significant (***) at the majority of stations and significance (**) at Sirt, Ajdabyia and Sabha (Table 4).The analysis of TXn indicated low positive changes at Tripoli Airport, Nalut, Misurata, Binina, Shaht and Darnah, with no changes at the remaining stations (Figure 6).Positive significant changes in TN10p are identified at only five stations.The decreases in TN10p are identified at Zwarah, Shahat, Al-Jaghbub (**) and Al-Kufrah (**), ranging between -0.077 and -0.033 day/year, with no changes at the remaining stations (Figure 7).Positive trends in TX10p are found at Tripoli Airport (0.037 day/year) and Al-Jaghbub (0.043 day/year), with high significance (**).In addition, a positive trend is found at Ajdabyia (0.001 daya-1), with low significant trends (*) and no changes at the remaining stations (Table 3 and Figure 8).

Conclusion
The results indicated that a significant increase in the coldest extreme (TNn) temperature was found at the majority of stations across the country.A mixture of increasing and decreasing trends in warmest day (TXx) are found, with significant changes more evident in autumn and spring than in winter and summer.The results of extreme temperature identified a significant increasing trend in extreme cold temperature in the monthly minimum value of daily minimum temperature (TNn) at all stations.A mixture of decreases and increases in maximum extreme temperature trends in the monthly maximum value of daily maximum temperature (TXx) are also found.Therefore, regional and local temperature trends can be strongly influenced by regional variability and changes in the climate system.The results of extreme temperatures identified that Libya is more affected by warming in the coastal region.However, the inland region was more affected by warming than the coastal region.Moreover, the minimum temperature showed that there were increasing trends for the mean average temperature at all stations across Libya.Increases in cold temperature extremes are smaller than those found for mean minimum temperature at most stations.Temperature variability can be associated with variations in large-scale atmospheric patterns represented by Eastern Atlantic and Western Mediterranean Oscillations, resulting from increases in atmospheric circulation, which seems to play a significant role in explaining the spatial and temporal variability of temperatures in the Mediterranean basin.Some regional differences between the model and observations may be due to local forcing or changes in climate dynamics.

Fig. 1 :
Fig. 1: Desiccation and death of trees (Cupressus sempervirens).Photo taken in Janzour city, June 2021 .2. Literature ReviewThe Fifth Assessment Report of the IPCC[9]  indicated that the surface temperature showed a change tendency rate of increase from 1880 to 2012, with an increase of 0.85 °C[9].Studies of observational temperature records over the last 50-100 years have found evidence for an increase in both observed mean, minimum and maximum temperature[10,5].In the global scope, the highest trend of the lowest temperature indices, the increase in the extreme temperature indices and the decrease in the cold extreme value all show obvious warming trends from the 20th century to the beginning of the 21st century[1,10].Morak et al.[11]  indicated an increase in warm extremes and a decrease in cold extremes in both seasons and in almost all regions that are generally well captured by the model.Ageena indicated that a mix of increasing and decreasing trends on the warmest day are also found.Zhou et al.[13] showed that the extreme temperature warming indices and the minimum daily maximum temperature (TXn) and the minimum daily minimum temperature (TNn) of cold indices showed an increasing trend from 1960 to 2016, especially since the 1990s.Rodríguez et al.[14] found a significant trend of more warm days and fewer cold nights.Moreover, an increase in the number of heat waves has also been observed in southwestern North America (Tebaldi et al.[15].Shen et al.[16]  found an upwards tendency in the occurrence of warm days (TX90) and nights (TN90) and in the temperatures of the coldest day (TXn), coldest night (TNn), warmest day (TXx), and warmest night (TNx) in China and most climate regions.In most extreme temperature indices in Asia, the number of warm days and nights is increasing, and the increasing speed of the latter is significantly higher than that of the former[ 7  1  ].In the past 60 years in Central Asia and North America, the growth rate of the highest temperature (TMAX) was faster than that of the lowest temperature (TMIN)[11],[19].The rise rate of the lowest temperature in Iran is twice that of the highest temperature.Therefore, the frequency of extremely high temperatures increases and the frequency of extremely low temperatures decreases[20, [10, 14].Toreti and Desiato[21] confirmed that most of the indices, averaged

Table 2 : Trend of annual extreme maximum temperature with statistically significant levels of stations across Libya (1961-2020)
Vol.22 No. 2 2023 30