Eswatini

The predictability of future tropical cyclones is marked with significant uncertainties due to several factors. These include disparities among climate models, the inherent complexity of processes incorporated into Tropical Cyclone Models, and regional variations in cyclone formation, behavior, and dispersal. Moreover, Tropical Cyclone Models are primarily calibrated for current climate conditions, which may introduce additional biases when applied to future scenarios. In summary, the complex and often conflicting interactions among ocean temperatures, wind patterns, and atmospheric conditions that drive cyclone formation, dispersal, and landfall remain poorly understood, making it challenging to identify which trends will dominate. For additional information on current understanding of tropical cyclones and tropical cyclone frequency, see Sobel et al. (2021). However, a few certainties do emerge in the projections. The clear trend of warming indicates that, once storms form and conditions become favorable, the intensities of these storms are likely to increase as compared to today. This signal is particularly pronounced for major tropical cyclones, as highlighted in the IPCC AR6 WG1 Report.

This page uses projections from the Columbia HAZard Model (CHAZ, Lee et al. 2018 ). These projections are derived from outputs of 12 CMIP6 GCMs under the SSP2-4.5 Scenario, focusing on the period 2035-2064 (centered on 2050). It’s important to note that the CHAZ model does not include projections for cyclone activity in the South Atlantic, based on the assumption that low cyclone activity will continue in the near future. Another important caveat is that CHAZ uses two configurations to represent moisture: the model with column relative humidity (CRH), used here, projects an increasing tropical cyclone frequency in the future, while the model using saturation deficit (SD) projects a decrease. In the Western North Pacific, cyclone frequency scales linearly with the rise in global mean surface temperature, highlighting a direct link to anthropogenic greenhouse gas radiative forcing. In the Atlantic, the SD experiments show the same trend, but the CRH response is more complex and nonlinear, likely due to the higher sensitivity of cyclones to aerosol forcing compared to greenhouse gas forcing. However, in both configurations, the percentage of intense cyclones is increasing, making this a consistent and reliable trend. For a review of tropical cyclone models, see Knutson et al. (2020).

Currently CCKP presents a single cyclone model, configuration, and scenario for the initial presentation of this complex subject matter.  

CCKP is grateful for the expertise and guidance of the Columbia HAZard Model (CHAZ) Team at Columbia University. 

 

01 - Projected Tropical Cyclones Activity Maps

Below, you will find the projected exceedance probability distribution of tropical cyclones (number of cyclones per year) and the corresponding return period (pull-down menu). By dragging the slider to the left, you can view the projected fractional change by 2050 compared to the historical period for cyclones of at least the selected category. 

What you can see in this figure
The map on the left shows the projected annual exceedance probability, representing the number of cyclones per year of at least the selected category in each 0.5° pixel projected for the future period, with levels ranging from blue (less frequent/less likely) to red (more frequent/more likely) as in the historical tab. Alternatively, the user can also select to visualize the projected return period. The return period is the inverse of the exceedance probability and indicates the expected average time between storms of a given category or higher. A higher return period indicates a lower frequency (or probability) of storms, and vice versa.

In comparison, the map on the right illustrates the fractional change between the historical and projected periods. A value of <1 indicates a decrease in the probability of storms (or increase in return period), while >1 indicates an increase in storm probability (or decrease in return period). Data presentations on this page represent the modeled uncertainties by displaying the inter-model projected changes in the 10th percentile, median, and 90th percentile in the anomaly plots (on the right). 

Understanding the Data: Implications and Utility
Note the direction of change and the magnitude, but always noting that this is just one model and one configuration. 

What are the key caveats and limitations to consider?
Projected Tropical Cyclone(TC) outcomes presented here are based on the same bias correction ratio as was applied to the historical model outcomes vis-a-vis observations. Spatial detail of these projections, and especially with regard to their change, is to be interpreted with caution as the class of models used for TC simulation remains simplified. The track of each storm was traced in space, including over land areas, to get these maps. In contrast to spatially aggregated landfall products (see table and graphs below), no separate bias-correction was applied in the map to landfall events and thus counts over land might be exaggerated. Note, the largest anomalies appear in areas where the counts are approaching very small numbers (close to the equator, and in the farthest reaches of TC tracks over land where dissipation is rapid). 

02 - Comparison of Historical and Projected Cyclone Statistics

In the charts below, we compare the historical and projected simulated percentages and counts of cyclones across various categories. These comparisons are made at the global level, ocean basin level, Exclusive Economic Zone (EEZ), and at the country or territory level associated with the EEZ, reflecting the number of landfalls.

By default, the data is presented at the ocean basin level and for the primary EEZ linked to the selected country. However, users can choose other EEZs from the same country, as well as select additional EEZs from the map above. Since some EEZs span multiple basins, we default to show the primary ocean basin, but users can select additional basins that intersect with that EEZ as needed.​

What you can see in this figure
Here, we calculate the annual total number of cyclones globally for each category with the number of cyclones within the selected ocean basin, Exclusive Economic Zone (EEZ), and associated landfall territory during both the historical and projected future periods. We present the projected fractional change, calculated as the ratio between the future and the present values. Values below 1 indicate a reduction in frequency, and values above 1 suggest an increase in frequency projected for the future. For instance, a value of 1.3 represents a 30% increase. 

Understanding the Data: Implications and Utility 
Notably, in most cases, the fractional changes for high-intensity tropical cyclones are more pronounced than those for weaker tropical storms.

What you can see in this figure
Here, we compare the percentage contribution of each category level for the global average, the average over the selected ocean basin, the average over the selected EEZ, and the distribution after cyclones make landfall (when they intersect with land), to a total of 100% for the historical (left columns) and projected scenarios (right columns).

The calculation process is as follows: Each cyclone is counted once for the global average and for each ocean basin. Cyclones are categorized based on their maximum wind speed, with each cyclone assigned to the ocean basin where it achieved its peak wind speed. For each EEZ and country (or territory), cyclones are counted and categorized according to the maximum wind speed within that specific region (EEZ or land boundaries). Note that a cyclone may reach Category 5 within an EEZ but weaken to Category 3 upon making landfall. In this case, the storm is counted as a Category 5 cyclone for the EEZ, but as a Category 3 cyclone for the corresponding country or territory. Similarly, a cyclone might reach Category 5 in an EEZ and weaken to Category 3 in another EEZ. In that case, the same cyclone will be counted as Category 5 in the first EEZ, and as Category 3 in the second EEZ.

Then, we calculate the percentage of each cyclone type. 

Understanding the Data: Implications and Utility 
Note how the proportion of low-category cyclones increases as the storm approaches land in most cases (from basin to EEZ to landfall). Moreover, the proportion of intense cyclones tends to increase in the future (not always) while the proportion of weaker tropical storms decreases. 

What you can see in this table
Here, we compare the annual total number of cyclones globally for each category with the number of cyclones within the selected ocean basin, Exclusive Economic Zone (EEZ), and associated landfall territory during both the historical and projected future periods. Additionally, we present the projected fractional change, calculated as the ratio between these values. Values below 1 indicate a reduction in frequency, and values above 1 suggest an increase in frequency projected for the future. For instance, a value of 1.3 represents a 30% increase. 

The calculation process is as follows: Each cyclone is counted once for the global average and for each ocean basin. Cyclones are categorized based on their maximum wind speed, with each cyclone assigned to the ocean basin where it achieved its peak wind speed. For each EEZ and country (or territory), cyclones are counted and categorized according to the maximum wind speed within that specific region (EEZ or land boundaries). Note that a cyclone may reach Category 5 within an EEZ but weaken to Category 3 upon making landfall. In this case, the storm is counted as a Category 5 cyclone for the EEZ, but as a Category 3 cyclone for the corresponding country or territory. Similarly, a cyclone might reach Category 5 in an EEZ and weaken to Category 3 in another EEZ. In that case, the same cyclone will be counted as Category 5 in the first EEZ, and as Category 3 in the second EEZ.