Hydrology of the Northern Areas

The streams and rivers of this area are very important to the overall hydrology of Pakistan since many of them flow into the Indus river or its tributaries. In fact the rivers of the Northern Areas contribute almost 72% of the mean annual flow of the Indus river (Government of Pakistan & IUCN, 2003). Stream flow measurements here are carried out by WAPDA’s Surface Water Hydrology Project and stretch back to 1960 (WAPDA, Lahore, pers. comm.).

Any river derives its flow ultimately from precipitation. But the storage of precipitation in the catchment has a great influence on the time distribution and magnitude of river flow. In this region the most important storage of water is in the glaciers and in the perennial and seasonal snow packs. Thus flow in the rivers and their peaks depend on the combined availability of heat to melt the snow and of water stored in the form of snow and ice.

Archer (2001) has calculated coefficients of variation of monthly flow in various rivers. This data gives an idea of the regularity and predictability of flows at each station. His calculations show that:

• basins that are mostly dependent on snow and glacial melt, such as Shyok, Shigar, Hunza and upper Indus, have lower variability in their flows while those which depend on rainfall in the southern part of the region, such as Gilgit, Brandu and Indus at Besham and at Shatial Bridge, have much higher variability.

• such variability is also smaller with increasing catchment size – thus Indus at Kachura has a more variable flow than the Indus at Besham.

• variability in flow is low during the winter months when most precipitation is bound up in snow and ice but when snow at lower altitudes begins to melt in March/April the flow increases. May is the month of the greatest variability. This seems to be due to the variable snow pack in the lower heights from year to year. In June higher snow fields and glaciers begin their annual melt but the variability in flow is less because in these heights the amount of snow and ice available for melting is more or less consistent from year to year.

Seasonality of annual maximum and minimum discharge shows that the higher catchments of Shyok, Shigar and Hunza have the latest mean maximum discharge dates from late July into August. This is because they are primarily fed by higher altitude snow and ice. Moving downstream, the date of maximum discharge becomes earlier (in June) such as at Astore and Kunhar.

Relationship of climate with hydrological regimes

A good analysis of the climatic influence on hydrological regimes in the Northern Areas is given in Archer (2003, 2001). Based on the relationships between seasonal climate and runoff he has suggested that the Upper Indus basin can be divided into three hydrological regimes.
High altitude catchments with large glacierized parts (e.g., Hunza and Shyok) with summer runoff that is strongly dependent on concurrent energy input represented by temperature.
Middle altitude catchments south of the Karakoram (e.g., Astore and Kunhar) that have summer flow mostly dependent on preceding winter precipitation.
Foothill catchments (e.g., Khan Khwar) that have a runoff regime controlled mainly by current liquid precipitation, predominantly in winter but also during the monsoon.

Archer (2001) says that summer runoff on the high altitude glacier-fed catchments is positively correlated with summer temperatures. He suggests a 17% increase in summer runoff for Shyok for 1oC temperature rise. However, runoff and temperature are negatively correlated on middle altitude snow-fed catchments. The reason is that increased temperature results in increased evaporation and, since the volume of snow cover is limiting, in reduced runoff, with an estimated reduction of almost 18% for a 2oC rise in temperature.

Summer runoff on the snow-fed catchments is positively correlated with winter precipitation. An increase in runoff of 5% at Astore and 3% at Kharmong for a 10% increase in mean winter precipitation has been suggested by the author. However, no significant correlation was found for the predominantly glacier-fed catchments. On the foothill catchments, significant correlation was found with spring (April to June) but not with summer runoff.

Such a variety of runoff responses to changes in the climatic variables means that it would be complicated to predict runoff response to climate change, and care should be taken to consider all these various responses in assessing the impact of variable climate on the hydrology of the region. This also has implications for water management in the area and downstream in the Indus.