Use of machine learning techniques for modeling of snow depth
Snow exerts significant regulating effect on the land hydrological cycle since it controls intensity of heat and water exchange between the soil-vegetative cover and the atmosphere. Estimating of a spring flood runoff or a rain-flood on mountainous rivers requires understanding of the snow cover dyn...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | Russian |
| Published: |
Nauka
2017-04-01
|
| Series: | Лëд и снег |
| Subjects: | |
| Online Access: | https://ice-snow.igras.ru/jour/article/view/357 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849338740805206016 |
|---|---|
| author | G. V. Ayzel |
| author_facet | G. V. Ayzel |
| author_sort | G. V. Ayzel |
| collection | DOAJ |
| description | Snow exerts significant regulating effect on the land hydrological cycle since it controls intensity of heat and water exchange between the soil-vegetative cover and the atmosphere. Estimating of a spring flood runoff or a rain-flood on mountainous rivers requires understanding of the snow cover dynamics on a watershed. In our work, solving a problem of the snow cover depth modeling is based on both available databases of hydro-meteorological observations and easily accessible scientific software that allows complete reproduction of investigation results and further development of this theme by scientific community. In this research we used the daily observational data on the snow cover and surface meteorological parameters, obtained at three stations situated in different geographical regions: Col de Porte (France), Sodankyla (Finland), and Snoquamie Pass (USA).Statistical modeling of the snow cover depth is based on a complex of freely distributed the present-day machine learning models: Decision Trees, Adaptive Boosting, Gradient Boosting. It is demonstrated that use of combination of modern machine learning methods with available meteorological data provides the good accuracy of the snow cover modeling. The best results of snow cover depth modeling for every investigated site were obtained by the ensemble method of gradient boosting above decision trees – this model reproduces well both, the periods of snow cover accumulation and its melting. The purposeful character of learning process for models of the gradient boosting type, their ensemble character, and use of combined redundancy of a test sample in learning procedure makes this type of models a good and sustainable research tool. The results obtained can be used for estimating the snow cover characteristics for river basins where hydro-meteorological information is absent or insufficient. |
| format | Article |
| id | doaj-art-1c72b18c3ca84f3bac947b34c04c839a |
| institution | Kabale University |
| issn | 2076-6734 2412-3765 |
| language | Russian |
| publishDate | 2017-04-01 |
| publisher | Nauka |
| record_format | Article |
| series | Лëд и снег |
| spelling | doaj-art-1c72b18c3ca84f3bac947b34c04c839a2025-08-20T03:44:18ZrusNaukaЛëд и снег2076-67342412-37652017-04-01571344410.15356/2076-6734-2017-1-34-44302Use of machine learning techniques for modeling of snow depthG. V. Ayzel0Institute of Water Problems, Russian Academy of SciencesSnow exerts significant regulating effect on the land hydrological cycle since it controls intensity of heat and water exchange between the soil-vegetative cover and the atmosphere. Estimating of a spring flood runoff or a rain-flood on mountainous rivers requires understanding of the snow cover dynamics on a watershed. In our work, solving a problem of the snow cover depth modeling is based on both available databases of hydro-meteorological observations and easily accessible scientific software that allows complete reproduction of investigation results and further development of this theme by scientific community. In this research we used the daily observational data on the snow cover and surface meteorological parameters, obtained at three stations situated in different geographical regions: Col de Porte (France), Sodankyla (Finland), and Snoquamie Pass (USA).Statistical modeling of the snow cover depth is based on a complex of freely distributed the present-day machine learning models: Decision Trees, Adaptive Boosting, Gradient Boosting. It is demonstrated that use of combination of modern machine learning methods with available meteorological data provides the good accuracy of the snow cover modeling. The best results of snow cover depth modeling for every investigated site were obtained by the ensemble method of gradient boosting above decision trees – this model reproduces well both, the periods of snow cover accumulation and its melting. The purposeful character of learning process for models of the gradient boosting type, their ensemble character, and use of combined redundancy of a test sample in learning procedure makes this type of models a good and sustainable research tool. The results obtained can be used for estimating the snow cover characteristics for river basins where hydro-meteorological information is absent or insufficient.https://ice-snow.igras.ru/jour/article/view/357boostingmachine learningmodelingopen datasnow depth |
| spellingShingle | G. V. Ayzel Use of machine learning techniques for modeling of snow depth Лëд и снег boosting machine learning modeling open data snow depth |
| title | Use of machine learning techniques for modeling of snow depth |
| title_full | Use of machine learning techniques for modeling of snow depth |
| title_fullStr | Use of machine learning techniques for modeling of snow depth |
| title_full_unstemmed | Use of machine learning techniques for modeling of snow depth |
| title_short | Use of machine learning techniques for modeling of snow depth |
| title_sort | use of machine learning techniques for modeling of snow depth |
| topic | boosting machine learning modeling open data snow depth |
| url | https://ice-snow.igras.ru/jour/article/view/357 |
| work_keys_str_mv | AT gvayzel useofmachinelearningtechniquesformodelingofsnowdepth |