3.3 Input data
Daily air temperature and humidity, and daily precipitation totals from meteorological stations of the hydrometeorological network within or nearby the basin were used as meteorological inputs for hydrological modelling. Four of them, Suntar-Khayata, Nizhnyaya Baza, Vostochnaya and Agayakan, were used for the period 1957-1964 and two, Agayakan and Vostochnaya, for the period 1966-2012 (Table 3, Fig. 1). Data from meteorological stations are interpolated to each RP. The interpolation is based on the triangulation method, when ideally each RP is inside a triangle, in the corners of which there are weather stations. Linear interpolation is conducted between the stations if only two are available.
The study area is characterized by temperature inversions. Annual average monthly temperature and air saturation deficit lapse rates were estimated using the data from the Suntar-Khayata and Agayakan meteorological stations (the range of elevation is 1292 m), they change from +1.1 ºС and +0.01 mbar per 100 m elevation increase in January to -1.3 ºС and -0.35 mbar per 100 m in June. The estimated values were used to correct interpolated temperature and saturation deficit from meteorological stations to RPs depending on the difference in elevation.
Data from four meteorological stations (Suntar-Khayata, Nizhnyaya Baza, Vostochnaya and Agayakan) from Reference book (1968) and the information of snow surveys at high mountain elevation (Grave, 1960) were used to analyze the distribution of precipitation at different altitudes for warm (May – August) and cold (September – April) periods of the year.
Annual precipitation at the Suntar-Khayata Station exceeds the precipitation amount observed at the foothills by more than twofold. The precipitation gradient for the altitude range 777 to 1350 m a.s.l. is 7 mm (5-7%) per 100 m, and at the altitude range 1350 to 2068 m a.s.l. it exceeds 35 mm (15-16%). Snow survey data for 1957-1959 (Grave, 1960) has demonstrated that altitudinal gradients of precipitation increase are steady and equal on average to 35 (5-8%) and 30 (4-5%) mm per 100 m for the altitude ranges of 2068-2257 and 2257-2477 m a.s.l. correspondingly.
Solid precipitation share at 777 m a.s.l. is approximately 25% of the annual total, and at 2068 m a.s.l. it increases to 60%. Mean annual precipitation from 1957 to 1964 at the Suntar-Khayata Station is 555 mm.
Correct estimation of precipitation is difficult in mountainous areas wheresignificant biases occur especially for winter precipitation because of the effect of wind on snowfall (Groisman et al., 2014). There are several methods for precipitation corrections. They are mainly based on the coefficient on wind speed and wind protection, air temperature and precipitation type (WMO Report no. 67, 1998; Yang & Goodison, 1995). In Reference Book (1968) some adjustments are recommended for wind underestimation and wetting loss, which can reach up to 1.7 times (1.6 on average in cold season) for solid precipitation, and 1.3 times (1.16 times on average in warm season) for liquid precipitation, which leads to the annual precipitation amounts of 688 mm at 2068 m a.s.l. (Reference Book, 1968), and 800 mm at the mountain peaks (Vasiliev & Torgovkin, 2002).
Corrected values of precipitation at meteorological stations were used to develop the relationships between both liquid and solid precipitation amount and terrain elevation in the basin. Precipitation amount for each RP is assessed according to those relationships based on elevation and interpolated daily solid and liquid sums of precipitation are normalized.
4. Results
We used available observational data from the Suntar-Khayata Station to verify the model parameterization for the goltsy landscape.