9. Discussion and conclusion
The ongoing increase of air temperatures, changes in precipitation and permafrost degradation affect the hydrological cycle via seasonal redistribution of water balance elements, changes in soil wetness and ALT, intensification of ground and surface water runoff exchange (Makarieva, 2019a; Rawlins et al., 2010; Shiklomanov & Lammers, 2013; Tananaev et al., 2016; Walvoord & Kurylyk, 2016).
However, the research is complicated by the lack of observed data and the inaccessibility of the study area of the most of the Arctic (Bennett et al., 2015). In the conditions of insufficient information, typical for the high-altitude mountain permafrost regions, the main tool for studying the processes is the method of mathematical modeling. The Arctic places increased demands on hydrological models. The vast majority of hydrological models, well-established in areas with a temperate climate, cannot be used in the permafrost zone. The main requirements are the physical validity of models to natural processes, their versatility in terms of use, both in different landscapes and spatial and temporal scales, and most importantly, the ability to assess the parameters of the model based on the measured properties of landscapes.
Stationary observations in small research catchments are the main source of information on the physical mechanisms of runoff formation and changes of hydrological cycle. But such observations are often expensive to maintain (especially in hard-to-reach regions), require large number of specialists, and are difficult to perform without additional support from the state or other interested stakeholders. In these conditions, the information for improving the models can also be integrated from short-term studies at certain landscapes and watersheds. The concept of such studies was proposed by Vinogradov (1988), he called it “nonstationary research watersheds and plots”. The idea was further developed by Vinogradova and Vinogradov (2014) and Gartsman and Shamov (2015) who called this approach “mobile watershed”. According to this approach, one or several indicative watersheds or plots are selected in the study area. The research visits are short-term (approximately from 2-3 weeks to 2 months) and last for several years. The observations are vaster and less detailed.
Conducting short-term intensive observations on specially selected representative watersheds, despite their fragmentation, allow formulating a general idea of ​​the conditions of runoff formation and hydrological phenomena of the territory in question, and most importantly, approximate quantitative assessment of the parameters of mathematical models of hydrological processes (Vinogradov & Vinogradova, 2014). However, not having in their basis the tasks of model development and parametrization, or refining existing methods for calculating the flow characteristics, such observations lose most of their value and do not justify the investment.
The three-years extensive observations at the Suntar-Khayata Station can be regarded as the good example of such approach. Among the others expeditions to remote regions we may mention historical studies at the Putorano Plateau in 1988-1990 (Reports…, 1988-1990), recent studies at Chukotka (Tregubov et al., 2020) and the Lena River delta (Tarbeeva et al., 2020). We emphasize the need for open access to the detailed hydrometeorological data of such research sites which can provide the opportunity for multi-criteria assessment of hydrological models in different conditions of permafrost zone (e.g. Fang et al., 2018; Makarieva 2017, 2018a; Rasouli et al., 2019).
In this research, based on the observation data at the high altitude Suntar-Khayata Station (the Indigirka river upper reaches) under the program of the International Geophysical Year in 1957-1959, the parameters of the hydrological model Hydrograph were developed, which describe the runoff formation processes in the high mountain goltsy landscape of the Suntar river basin. Variable states of snow cover and heat dynamics in ground profile in the goltsy zone were simulated, as well as runoff formation process and its changes in recent time throughout the whole catchment of the Suntar river. Modeling results are considered acceptable.
Model calculations have allowed evaluation of long-term average annual values of water balance for different landscapes, their contribution to runoff formation in the mountain river outlet. It is established that the runoff formation occurs mainly in the high-altitude region of goltsy and tundra (about 70%).
The model reproduced observed trend values and change point of streamflow during the study period supporting the hypothesis that the increase of liquid precipitation in autumn due to climate warming could be the main factor of streamflow changes in the autumn-winter period in this permafrost region.
Currently, in the mountain regions of the Yana, Indigirka and Kolyma rivers basins no hydrological research stations are left to perform a comprehensive study of runoff formation processes. Therefore, development and verification of methods for hydrological processes modeling which may successfully utilize short-term, extremely scarce data, become more of great current interest. The presented study has demonstrated that the Hydrograph model and its further development may become a foundation for solving scientific and practical issues in the research region.