Page 328 - Hydrogeology Principles and Practice

HYDC08 12/5/05 5:32 PM Page 311

Groundwater resources and environmental management 311

BO X
Assessment of drought severity
8.7

Droughts and floods are characterized by the extremes of the fre- tal drought measures the significance of the drought for those
quency, intensity and amounts of precipitation (Trenberth et al. water uses directly affected by a shortage of precipitation, for ex-
2003). Interestingly, there is no single definition of drought. Beran ample aquatic ecology, fisheries, low river flows, reduced spring
and Rodier (1985) considered, without specifying when a period of flows and groundwater levels, agriculture and horticulture. A water
dry weather becomes a drought, that the chief characteristic of a supply drought measures the significance for those indirectly
drought is a decrease of water availability in a particular period over affected, for example the risk of water demand reductions imposed
a particular area. Meteorological drought is defined in terms of a on domestic and industrial water consumers. In many cases, a
deficit of precipitation. Agricultural drought relates mostly to drought will be classified as having both environmental and water
deficiency of soil moisture, while hydrological drought relates to supply impacts (Mawdsley et al. 1994).
deficiencies in, for example, lake levels, river flows and groundwater The communication of the severity of a drought can be simply
levels. The greatest severity and extent of drought in the United presented according to its intensity and duration and classified as
States occurred during the Dust Bowl years of the 1930s, particu- moderate, serious or severe (Table 1). The duration of an environ-
larly during 1934 and 1936. The decades of the 1950s and 1960s mental drought is of particular relevance to groundwater resources.
were also characterized by episodes of widespread, severe drought, If a region is largely dependent on groundwater, and if the available
while the 1970s and 1980s as a whole were unusually wet. aquifer storage is relatively insensitive to droughts only lasting
Drought conditions that began in the west of the United States in one recharge period, then this situation is likely to result in a short
1998 persisted in many areas through to the summer of 2003 environmental drought without significant implications for water
(Trenberth et al. 2004). In the United Kingdom, long time-series supplies and the aquatic environment. However, if the drought is
data of daily precipitation and monthly groundwater level provide of long duration during which there is a reduction in seasonal
an indicator of hydrological drought periods, as illustrated by com- recharge intensity, then a moderate drought may result with sig-
paring the Central England annual precipitation record and the nificant implications for water supplies, water levels and river flows
mean annual Chalk groundwater level for Chilgrove House in south- dependent on groundwater.
ern England (Fig. 1). A quantitative measure of environmental droughts is to calculate
Globally, drought areas increased more than 50% throughout a drought severity index. In the United States, the Palmer Drought
the twentieth century, largely due to the drought conditions over Severity Index (PDSI), as devised by Palmer (1965), is used to repres-
the Sahel and Southern Africa during the latter part of the century, ent the severity of dry and wet periods based on weekly or monthly
while changes in wet areas were relatively small (Trenberth et al. temperature and precipitation data as well as the soil water holding
2004). The most significant cause of drought worldwide is the El capacity at a location (Dai et al. 1998). Areas experiencing a severe
Niño Southern Oscillation (ENSO) which also emphasizes the con- drought score a PDSI of −4.0 while areas with severe moisture sur-
current nature of floods and droughts, with droughts favoured in plus score +4.0. Between these two extremes, 11 categories of wet
some areas during an ENSO event, while wet areas are favoured in and dry conditions are defined.
others. These areas tend to switch during La Niña in the tropics and A more generally applicable drought severity index (DSI), which
subtropics. Palaeoclimatological studies show evidence of dramatic may be considered for different hydrological data, is to calculate an
changes in drought and the hydrological cycle over many parts of accumulated monthly deficit relative to the mean for a standard
the world, with droughts lasting several decades not uncommon period (Bryant et al. 1992). In this approach, based on available
(Box 5.1). The full range of drought variability is probably much long-term hydrological records, a drought is considered to end
larger than has been experienced in the last 100 years (Trenberth when, for example, the 3-monthly precipitation total exceeds the 3-
et al. 2004). monthly mean for these months. It must be noted that the choice of
The shortage of water experienced by water users and the termination criterion must be considered carefully since different
aquatic environment during a drought is a complex issue. Poten- rules can produce different impressions of drought severity that are
tially, the full characterization of events that vary regionally and potentially inappropriate for some long-duration events. A disad-
temporally can involve the assessment of a wide range of hydrolo- vantage of this approach is that a DSI based only on precipitation
gical indices (for example precipitation amount, river flows, surface data does not directly indicate the impact on the environment since
reservoir storage and groundwater levels). According to Mawdsley this will depend on the antecedent soil moisture conditions in a
et al. (1994), there are essentially two types of users impacted catchment. Instead, drought severity may be better determined
by the shortage of water. Firstly, there are those who are affected using effective precipitation, or selected local river flow and ground-
directly by a deficit of precipitation, possibly compounded by high water level data, even though such records may only be available
evaporation, and secondly, those consumers who are affected indir- for a short time relative to precipitation records. Potentially,
ectly due to the way in which water storage facilities are managed effective precipitation (the balance between precipitation and
during a drought. Following from this, and including environmental evapotranspiration, eq. 8.2) could be a useful indicator of environ-
water requirements, it is recommended that droughts are classified mental drought, particularly in regions with significant ground-
either as environmental or water supply droughts. An environmen- water resources.