RECONSTRUCTION AND PREDICTION OF WATER BALANCE COMPONENTS FROM DENDROCHRONOLOGICAL DATA FOR THE NARYN RIVER BASIN (KYRGYZSTAN)

Abstract

Suggested new method for reconstruction of runoff includes: 1. Search informative points based weather observations for dependence runoff = f(climate index). Climate index consists of normalized anomalies of annual precipitation and air temperature, together with water vapor pressure and total cloud amount in the June–September. 2. The selection of sites that represent local relationship between the index of climate and tree rings. 3. Getting the multivariate linear regression equation between the runoff V and tree rings D. Basic climatic period 1961–1990 is used as a calibration interval of time. Verification of the equations for 1940–1960 identified the need to use other arguments in addition to the width of tree rings. For this purpose, two climate index: PDSI and SPEI were tested. It is established that the empirical formulae V = f(D, PDSI) and V = f(D, SPEI) are suitable for assessing the annual flow of the Naryn in 1901–2006. In the report is described a process of reconstruction of flow for the years 1700–2005 based on the data of the width of the tree rings, and taking into account the temporal variability of parameters of regression equations.Chronologies of the width and density of tree rings in the thirty-two tree ring sites were used to reconstruct long-term series of average summer air temperature Ts at meteorological stations of the Pamir and Tien Shan. The calibration interval of time was 1961–1990 to find equations Ts = f(D) of third-order multiple regression, and independent control of their quality, done in 1932–1960. Combined correlation coefficient in 10 cases out of 15 was more than 0.80, a relative error of calculation Ts in 19611990 ranged from 0.13 to 4.73%, and in 1932–1960 from 0.23 to 8.10%. Duration of the reconstructed series Ts ranged from 100 to 278 years. Positive results were received for predicting a common and glacial runoff of Naryn river, a long-term range of density of tree rings.

Keywords

• dendrochronology
• forecast
• hydroclimatological index
• multiple regressions
• tien shan
• water balance