The Abdi-Basid Courses Institute is a fully virtual platform aiming to unveil the engine of science and make it accessible to all. We offer online courses, live sessions, and multimedia content across fields like statistics, climate science, AI, and more. Our mission is to bridge science and society through inclusive, flexible, and engaging learning for students, professionals, educators, and curious minds worldwide.
This study investigates the atmospheric dispersion of volatile emissions from storage tanks in a petroleum depot using the AEROMET software. Emissions from tank facilities, including hydrocarbons and other volatile compounds, were estimated using standard EPA methodologies, with meteorological conditions processed through AERMET and topographic effects incorporated via AERMAP. The dispersion patterns were simulated to evaluate the spatial distribution of pollutant concentrations and identify potential areas of environmental and health risk within the depot zone.
(Jan, Feb, Mar, Apr, May, Jun, Jul Aug, Sep, Oct, Nov and Dec)
This research examines historical climatic variability and extreme weather risks in coastal environments using three state-of-the-art climate reanalysis products: ERA5, MERRA2, and CFSR. Covering the period 2000–2021, the analysis focuses on wind speed, precipitation, temperature, evapotranspiration, humidity, and sea level rise. Results highlight the dominance of moderate winds (2–3 m/s), with stronger extremes captured by MERRA2, while wind rose diagrams reveal prevailing easterly and southeasterly flows. Precipitation variability shows alternating phases of severe droughts (2000, 2008, 2012) and extreme wet anomalies (2003, 2010, 2019) identified through the Standardized Precipitation Index (SPI), underscoring the recurrent risks of droughts and flash floods. Seasonal contrasts in temperature and dew point reflect hot and humid summers versus cooler and moisture-rich periods, with evapotranspiration exhibiting an inverse relationship to precipitation and specific humidity peaking in late spring and early autumn. At a broader scale, sea level rise assessments indicate a global trend of +3.013 mm/year, with higher regional anomalies in adjacent ocean basins, while future projections under high-emission scenarios (RCP8.5) suggest increases exceeding +0.8 m by 2100. Overall, this comparative reanalysis provides an integrated assessment of coastal climatic variability and associated hazards, emphasizing the value of multi-source datasets for understanding regional climate risks and supporting adaptation strategies.
(a)(b)
Figure 1. Distribution of wind speeds (m/s) using (a) a histogram and (b) a box plot for the period 2000–2021.
Figure 1 illustrates the distribution of 10 m wind speeds from three satellite-based reanalysis products using a histogram (Figure 1a) and a box plot (Figure 1b). The histogram shows that wind speeds around 2.5 m/s are the most frequent, with a probability of at least 24%. The box plot (Figure 1b) provides additional statistics on the spread of wind speeds across the three datasets. A comparison indicates that the maximum wind speed in MERRA2 (12.49 m/s) is higher than in CFSR (8.15 m/s) and ERA5 (7.24 m/s). This discrepancy can be attributed to differences in spatial resolution, which is finer for CFSR and ERA5 (20–25 km) and coarser for MERRA2 (50–62.5 km). Overall, the dispersions in CFSR and ERA5 are slightly more similar to each other than to MERRA2.
Figure 2. Wind rose diagram at 10 m altitude over the period 2000–2021 from the CFSR, ERA5, and MERRA2 satellite products.
The wind rose diagram (Figure 2) illustrates the percentage distribution of wind by direction. As noted earlier, the 2–3 m/s wind speed class dominates the overall distribution. For all three satellite products (CFSR, ERA5, and MERRA2), the eastern sector is the most frequent in terms of wind speeds ranging from 1 to 6 m/s. The second most active sector is the southeast for ERA5 and MERRA2, while it is the northeast for CFSR. However, the CFSR product shows the highest wind speed class (7–8 m/s) in the southwest. For this same class, ERA5 indicates occurrences in the east and southeast sectors. In MERRA2, the highest wind speeds are recorded in the west and southwest sectors.
Overall, wind speeds are denser in the east, northeast, and southeast sectors, ranging between 1 and 6 m/s. In addition, the strongest winds (6–12 m/s, exceeding 43 km/h) are observed in the west and southwest sectors.
(a)(b)(c)
Figure 3. Monthly distribution of wind speeds (m/s) over the study period (2000–2021) from the ERA5, CFSR, and MERRA2 satellite products.
Based on the monthly variation from 2000 to 2021, Figure 3 shows a heterogeneous year-to-year variability in wind speeds. For the CFSR product, 2001 — particularly in July — experienced the highest wind speeds over the study period, while 2016 recorded relatively lower wind speeds. The diagram also indicates that wind speeds are generally stronger during the warm season (June to September) compared to the cooler season (October to May). For MERRA2, the highest wind speeds were observed in 2019 and 2021, with greater activity during the warm season than the cooler season. For ERA5, the year 2012 recorded the strongest wind speeds over the study area..
(a,b)
Figure 4. (a) Annual variation of precipitation (mm) and (b) frequencies of droughts, light, and heavy rainfall from the ERA5, CFSR, and MERRA2 satellite products over the period 2000–2021.
Figure 4a presents the annual variation of cumulative precipitation over the study period. Peak annual precipitation occurred in 2006 for ERA5 (684.20 mm) and CFSR (362.23 mm), and in 2019 for MERRA2 (619.47 mm). Additionally, the highest daily precipitation was recorded at 116.82 mm (MERRA2, 2019), 115.69 mm (CFSR, 2020), and 72 mm (ERA5, 2018).
Figure 4b shows the frequencies of precipitation classified into three categories: no rain, light rain, and heavy rain. The frequency of no rainfall (P = 0) ranges from 0 to 6% across all three products. Light rain (less than 10 mm) occurs with a frequency between 20 and 52%. Finally, the frequency of heavy rain (greater than 10 mm) is estimated between 42% and 80%, depending on the dataset.
(a,b)
Figure 4 presents the variation of drought and flood events assessed using the 3-month SPI over the period 2000–2021. Table 2 provides the classification of dry and wet events according to SPI values. The MERRA2 product indicates extremely wet years in 2003, 2005, and 2019 (Figure 5a). Similarly, ERA5 and CFSR identify 2019 as a very wet year (Figure 5a). Figure 5b highlights only the extreme events that can lead to human losses and environmental damage. Extremely dry periods were observed in 2000, 2002, 2008, 2009, and 2012 across all three products, while extremely wet periods were recorded in 2000, 2003, 2005, 2006, 2007, 2010, 2018, and 2019 (Figure 5b).
Figure 6. Monthly variation of maximum and minimum temperature (°C) from the ERA5, CFSR, and MERRA2 satellite products over the period 2000–2021.
The interannual variability of maximum and minimum temperatures for the period 2000–2021 is broadly similar across the three satellite products. The CFSR product appears to perform better than the others during the warm season. During this period, maximum temperatures generally peak between 33°C and 44°C, while they decrease during the cooler season from October to May.
Figure 7. Monthly variation of dew point temperature (°C) derived from the ERA5 and MERRA2 satellite products over the period 2000–2021.
To measure the level of humidity in the air, the dew point can be used. The higher the dew point, the greater the amount of moisture in the atmosphere, which directly affects the level of outdoor comfort. The different dew point ranges according to comfort categories are provided in Table 3. During the warm season, dew point temperatures are lower than during the cool season. According to the classification of the Australian Government Bureau of Meteorology, the warm season in MERRA2 corresponds closely to the 15–20 °C class (a combination of heat and humidity). In contrast, during the cool season, atmospheric humidity is higher (Figure 7).
Figure 8. Monthly variation of precipitation and evapotranspiration (mm) from the ERA5, CFSR, and MERRA2 satellite products over the period 2000–2021.
Figure 8 illustrates the interannual variation of evapotranspiration during the study period (2000–2021). Results show that for all three satellite products (CFSR, ERA5, and MERRA2), evapotranspiration increases during the warm season and decreases during the cool season. In addition, Figure 8 highlights that precipitation and evapotranspiration are inversely proportional throughout the months.
Figure 9. Monthly variation of specific humidity (g/kg) and mean temperature (°C) over the period 2000–2021.
Figure 9 presents the interannual variation of specific (or absolute) humidity, expressed in g/kg, derived from the MERRA2 and CFSR satellite products over the period 2000–2021. This parameter represents the number of grams of water vapor in a given volume, relative to the mass of dry air in that volume, expressed in kilograms. Both MERRA2 and CFSR indicate that in April and May, specific humidity reaches its highest values compared to the other months. In contrast, June and July record the lowest levels, while August and September show the second-highest period of specific humidity during 2000–2021. These results are consistent with the monthly variation of dew point temperature.
(a,b)
Figure 10. (a) Increase in sea level (mm) at global and regional scales between 1992 and 2021, and (b) location of the seas.
Figure 10a shows the annual mean variation of sea level rise at three scales: (i) global oceans, (ii) the Arabian Sea, and (iii) the Indian Ocean over the period 1992–2021. The locations of these regions are shown in Figure 10b. During this period, the global mean sea level rose significantly at a rate of 3.013 mm/year. The trend is slightly higher in the Arabian Sea, with 3.397 mm/year, while the Indian Ocean shows a rise of 3.115 mm/year.
Projections of global mean sea level rise for 2081–2100 relative to 1986–2005 are likely to be within the following ranges (given with confidence intervals):
+0.26 to +0.55 m for RCP2.6,
+0.32 to +0.63 m for RCP4.5,
+0.33 to +0.63 m for RCP6.0,
+0.45 to +0.82 m for RCP8.5.
For RCP8.5, the projected rise by 2100 is between +0.52 and +0.98 m, with an average annual rate of 8–16 mm during 2081–2100.
Description :This study investigates the climatic dynamics and meteorological variability of a coastal zone located in an arid environment and strongly influenced by the African monsoon and large-scale atmospheric circulations from the Indian and Pacific Oceans. Climate reanalysis datasets from ERA5 (ECMWF/C3S), CFSR (NCEP), and MERRA2 (NASA/GMAO) are employed to characterize long-term trends and climate extremes over the period 2000–2021. The analysis focuses on multiple parameters, including precipitation, temperature, humidity, wind, evapotranspiration, cloud cover, and sea level rise, with a spatio-temporal resolution tailored to strategic coastal environments.
The findings reveal pronounced interannual irregularities in precipitation, a significant intensification of hydrometeorological extremes (droughts and flash floods), contrasting seasonal wind regimes associated with monsoonal dynamics and the Khamsin phenomenon, as well as a progressive rise in sea level consistent with broader trends observed in the Indian Ocean. These insights contribute to a better understanding of regional climate risks and provide critical knowledge to support the resilience of coastal and port infrastructures in the area.
Umbrothermal diagram showing the monthly variation of precipitation (mm) and temperature (°C) from 2017 to 2021 of Djibouti airport station is presented in Figure 1.
The study area is marked by low rainfall. The variation in precipitation from 2017 to 2021 shows an irregular pattern. For example, the cumulative rainfall in 2019 was 427 mm. It is almost 3 times that of 2018 (138 mm), which in turn is almost 2 times that of 2017 (59mm). The total monthly rainfall in last year, recorded by the meteorological station at Djibouti-Ambouli airport, was 143.6 mm (Figure 1). The interannual variation (in-line curve) shows an irregular rainfall pattern with occasional heavy downpours that can cause flooding. In addition, the study area receives most of its rainfall in two seasons, March-May (also called long rains) and October-December (also called short rains). In particular, two consecutive years (2020 and 2021) show that April is wettest than the other months of the year.
Figure 1. Umbrothermal diagram showing monthly variation in rainfall (mm) and temperature (°C) from 2017 to 2021 for the Djibouti airport station which is the closest to the study area.
As far as temperature (°C) is concerned, the study area shows a more regular annual variation from 2017 to the 2021. The average temperature rises during the season from April to September (warm season) and falls from October to March (cool season). The peak is usually reached in July. The 2017 hot season was the warmest with an average temperature of 35.51°C. Similarly, the coolest season, with an average temperature of 27.35°C, was observed in the same year. Overall, it can be seen that precipitation is more abundant during the cool season than during the hot season (Figure 1).
Due to the unavailability of relative humidity data for the airport station, we use simulated data from ERA5. The yearly variation in relative humidity (%) from 2017 to 2021 is presented in Figure 2. Over the study period, the interannual variation in relative humidity follows a regular pattern from year to year. In particular, we can see in the figure 2, that the relative humidity is high in 2019 during the months of October, November and December with 77.79%, 78.68% and 81.70% respectively. During this season, the Djibouti airport station recorded a cumulative rainfall of 401mm, which caused human and material damage. In addition, in April 2020, the relative humidity reached 80.37% (figure 2). This coincides with a rainfall of 80mm with data from the airport station. On the other hand, the relative humidity is below 70% from June to September, which corresponds to a season of low rainfall with an increase in average temperature (Figures 1 et 2).
Figure 2. Monthly variation in relative humidity (%) between 2017 and 2021 in the study area.
Due to the unavailability of evapotranspiration data for the airport station, we use simulated data from ERA5. The yearly variation of potential evapotranspiration obtained with the Hargreaves method is presented in Figure 3. There is an almost regular annual variation from 2017 to 2021. During the warm season, the monthly evapotranspiration exceeds 90 mm. This indicates that the increase in temperature influences the evapotranspiration potential and leads to its highest value. On the other hand, during the cool season, which is also the season of dense rainfall, the evapotranspiration is at its lowest level of the year (Figure 3).
Figure 3. Monthly variation in potential evapotranspiration (mm) between 2017 and 2021 for the study area.
Due to the unavailability of cloud cover data for the airport station, we use simulated data from ERA5. The yearly variation in cloud cover over the study area during the period 2017 to 2021 is presented in Figure 4. An irregular pattern can be observed from year to year. In particular, during the months of February, March, April, July and November, the cloud cover exceeds 50% over the study area. Furthermore, three peaks (exceeding 55%) can be observed, namely in February 2017, November 2019 and April 2020. During these periods, heavy rainfall of 12 mm, 338 mm and 80 mm was recorded at the airport station. However, this is not always the case, for example, in February 2018, the airport station recorded no rainfall while the cloud cover was 52.48%.
Figure 4. Monthly variation in cloud cover between 2017 and 2021 for the study area.
Due to the unavailability of wind speed data for the airport station, we use simulated data from ERA5. Figure 5 shows the average variation in monthly wind speed occurred between 2017 and 2021. It can be seen that the wind speeds exceed 4 m/s in February, July and August. These are the windiest months of the year. In contrast, the wind speed is lower during the months of May, June, September and October.
Figure 5. Monthly variation of average wind speeds between 2017 and 2021 for the study area.
Due to the influence of the African monsoon, the wind comes from the east (maritime wind) during the month of September to May and from the west and southwest (Khamsin wind) between June and August. The wind rose of the study area at an altitude of 10 meters is presented in figure 6a. During the period 2007 to 2021, wind speeds are dominant in the east, northeast and southeast directions (from 45° to 135°) with a frequency of 15 to 25%. The highest wind speed class (which is also the least frequent) in this area is between 6 to 7 m/s. In contrast, to the north and south of the study area, the wind speed frequencies are less than 10% and vary between 1 and 7 m/s. In addition, the west and south-west direction (225° to 290°C) represents the area where wind speeds exceed 8 m/s. Overall, this region is particularly marked by the highest wind speeds in the study area, but with a frequency of less than 10%. On the other hand, it can be seen that hourly wind speeds above 5m/s were particularly active in 2020 and 2021. It can be said that this is the windiest year. Moreover, between 10 am and 3 pm, winds reach speeds above 3m/s between 2017 and 2021. In contrat, from 5 pm to 5 am, the wind speeds can be seen below than 3 m/s.
(a)
(b)
(c)
Figure 6. (a) Wind rose diagram and (b, c) Hourly and daily variation of wind speeds (m/s) for the year 2021.
Figure 7 shows the average annual change in global and Indian Ocean sea level rise over the period 1992 to 2021. During this period, global sea level has risen significantly by 3,013 mm/year. This trend is slightly higher in the Indian Ocean, which shows a sea level rise rate of about 3,115 mm/year. According to the IPCC climate panel prediction scenarios, with the RCP8.5 model, the global mean sea level rise in the years 2046 to 2065 is likely to be in the range of 0.22 to 0.38 m and for the years 2081 to 2100 it would be in the range of 0.45 to 0.82 m. These projected increases in sea level will not affect the current project in the sense that the Damerjogue oil terminal project is at an elevation of more than 3 meters. However, a possible sunnami cannot be excluded.
Figure 7. Change in sea level rise from 1992 to 2021.
Figure S1. Monthly variation of specific humidity (g/kg) and evapotranspiration (mm) during the year 2021.
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Figure S2. Seasonal distribution of wind directions and frequencies (%) from January to December (2017–2021).
