| ||UNESCO at Mauritius|
| ||BARBADOS +10|
| ||UNESCO & SIDS|
| ||LOOKING FORWARD|
| ||UNESCO.ORG is also:|
Many of these are correlated with El Niño-La Niña events. Limited water storages on small islands mean that domestic water supplies are severely threatened by these dry periods. In this work, meteorological or climatological drought is used to define drought relevant to domestic water supplies. South Tarawa, the capital of the Republic of Kiribati, is used as a case study. It is argued that drought depends on the characteristics of the water sources. Three sources of water are identified – rain tanks, domestic water wells and reticulated water supplied from large freshwater lenses and these need to be treated separately for drought declaration.
| ||Droughts in Small Coral Islands : Kiribati Case Study (1999)Droughts are frequent events in small coral islands and small island nations across the Indian and Pacific Oceans.|
Since the key components of the water balance are not known for rain tanks or domestic wells, a method of ranking rainfalls totalled over different periods was undertaken to identify and rank droughts. The Palmer Drought Index, the Standardized Precipitation Index, the Decile Method, and the Rainfall Depreciation Method are all examined critically. Here the decile method is explored and droughts are defined as periods when the rainfall drops below the 10 percentile level of all rainfalls for the totalling period in question. For Tarawa, it is suggested for meteorological drought that the appropriate totalling periods are: 4 to 6 months for rain tanks; 12 to 30 months for domestic wells; and 60 to 120 months for large freshwater lenses.
Using this method of ranking dry periods, it was found that severe droughts for rainwater tanks in South Tarawa occur on average approximately every 4 to 5 years and have an average duration of 5 to 6 months. For domestic wells, severe droughts occur on average 5.5 to 6 years and last, on average, 7 to 8 months. Because these events are frequent, policy can be developed to address response to drought. For the large freshwater lenses, drier periods occur on average every 6 years and last on average about 12 months. However, significant recharge of large freshwater lenses still occurs in these drier periods because of the long period over which rainfall is summed and the small variation in rainfalls over these long periods.
It was found that the Standardized Precipitation Index and the Rainfall Depreciation Method gave identical results to the Decile Method. The greater simplicity of the decile method and the ease of understanding of its rainfall rankings suggests that is has advantages over other methods. The appropriateness of the 10 percentile level as an indicator of severe drought and the 40 percentile level as a warning trigger for droughts, compared to other levels, remain to be examined in detail. Likewise, the appropriateness of the rainfall totalling periods suggested here also should be tested against experience for a range of small island nations and conditions. Data on the size of rainfall catchments, rain tanks and details of water usage from them and from domestic wells should be collected. While the methodologies used here are applicable elsewhere, the results and conclusions are specific to Betio, South Tarawa. It is therefore suggested that donor agencies could use this system of ranking rainfall as a method of determining support in dry periods for small island nations. To do this, appropriate rainfall statistics, information on water sources and testing of concepts will be required for small island nations.