Chalachew Muluken Liyew
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Chalachew Muluken Liyew has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
Elvira Di Nardo
3 shared publicationsRosa Meo
3 shared publications- Stefano Ferraris
Interuniversity Department of Regional and Urban Studies and Planning (DIST), University of Torino and Polytechnic University of Torino, 10125, Torino, Italy
3 shared publications - Elvira Di Nardo
Department of Mathematics “G. Peano”, University of Turin, 10123 Turin, Italy
2 shared publications Haileyesus Amsaya Melese
Bahir Dar University, Bahir Dar Institute of Technology, Bahir Dar, Ethiopia
2 shared publications- Rosa Meo
University of Turin
2 shared publications
Latest Publications
- Open Access
- 4 July 2025
- Published by Springer Nature in Artificial Intelligence Review
- Vol. 58 (10) , 1-37
A review of feature selection methods for actual evapotranspiration prediction
- 4 July 2025
- Published by Springer Nature in Artificial Intelligence Review
- Vol. 58 (10) , 1-37
Open AccessA review of feature selection methods for actual evapotranspiration prediction
- Open Access
- 6 May 2025
- Published by Elsevier in Machine Learning with Applications
- Vol. 20
Hyperparameter optimization of machine learning models for predicting actual evapotranspiration
- 6 May 2025
- Published by Elsevier in Machine Learning with Applications
- Vol. 20
Open AccessHyperparameter optimization of machine learning models for predicting actual evapotranspiration
- Open Access
- 15 November 2024
- Published by Copernicus GmbH in Advances in Statistical Climatology, Meteorology and Oceanography
- Vol. 10 (2) , 173-194
Identifying time patterns of highland and lowland air temperature trends in Italy and the UK across monthly and annual scales
This paper presents a statistical analysis of air temperature data from 32 stations in Italy and the UK up to 2000 m above sea level from 2002 to 2021. The data came from both highland and lowland areas in order to evaluate the differences due to both location and elevation. The analysis focused on detecting trends at annual and monthly timescales, employing ordinary least-squares (OLS), robust S-estimator regression, and Mann–Kendall (MK) and Sen's slope methods. Hierarchical clustering (HCA) using dynamic time warping (DTW) was then applied to the monthly data to analyze the intra-annual pattern similarity of trends within and across the groups. Two different regions of Europe were chosen because of the different climate and temperature trends – namely, the northern UK (smaller trends) and the northwest Italian Alps (larger trends). The main novelty of the work is to show that stations with similar locations and altitudes have similar monthly slopes by quantifying them using DTW and clustering. These results reveal the nonrandomness of different trends throughout the year and between different parts of Europe, with a modest influence of altitude in wintertime. The findings revealed that group average trends were close to the National Oceanic and Atmospheric Administration (NOAA) values for the areas in Italy and the UK, confirming the validity of analyzing a small number of stations. More interestingly, intra-annual patterns were detected commonly at the stations of each of the groups and are clearly different between them. Confirming the different climates, most highland and lowland stations in Italy exhibit statistically significant positive trends, while in the UK, both highland and lowland stations show statistically nonsignificant negative trends. Hierarchical clustering in combination with DTW showed consistent similarity between monthly patterns of means and trends within the group of stations and inconsistent similarity between patterns across groups. The use of the 12 distance correlation matrices (dcor) (one for each month) also contributes to what is the main result of the paper, which is to clearly show the different temporal patterns in relation to location and (in some months) altitude. The anomalous behaviors detected at 3 of the 32 stations, namely Valpelline, Fossano, and Aonoch Mòr, can be attributed, respectively, to the facts that Valpelline is the lowest-elevation station in its group; Fossano is the southernmost of the Italian stations, with some sublittoral influence; and Aonoch Mòr has a large number of missing values. In conclusion, these results improve our understanding of temperature spatio-temporal dynamics in two very different regions of Europe and emphasize the importance of consistent analysis of data to assess the ongoing effects of climate change. The intra-annual time patterns of temperature trends could also be compared with climate model results.
- 15 November 2024
- Published by Copernicus GmbH in Advances in Statistical Climatology, Meteorology and Oceanography
- Vol. 10 (2) , 173-194
Open AccessIdentifying time patterns of highland and lowland air temperature trends in Italy and the UK across monthly and annual scales
This paper presents a statistical analysis of air temperature data from 32 stations in Italy and the UK up to 2000 m above sea level from 2002 to 2021. The data came from both highland and lowland areas in order to evaluate the differences due to both location and elevation. The analysis focused on detecting trends at annual and monthly timescales, employing ordinary least-squares (OLS), robust S-estimator regression, and Mann–Kendall (MK) and Sen's slope methods. Hierarchical clustering (HCA) using dynamic time warping (DTW) was then applied to the monthly data to analyze the intra-annual pattern similarity of trends within and across the groups. Two different regions of Europe were chosen because of the different climate and temperature trends – namely, the northern UK (smaller trends) and the northwest Italian Alps (larger trends). The main novelty of the work is to show that stations with similar locations and altitudes have similar monthly slopes by quantifying them using DTW and clustering. These results reveal the nonrandomness of different trends throughout the year and between different parts of Europe, with a modest influence of altitude in wintertime. The findings revealed that group average trends were close to the National Oceanic and Atmospheric Administration (NOAA) values for the areas in Italy and the UK, confirming the validity of analyzing a small number of stations. More interestingly, intra-annual patterns were detected commonly at the stations of each of the groups and are clearly different between them. Confirming the different climates, most highland and lowland stations in Italy exhibit statistically significant positive trends, while in the UK, both highland and lowland stations show statistically nonsignificant negative trends. Hierarchical clustering in combination with DTW showed consistent similarity between monthly patterns of means and trends within the group of stations and inconsistent similarity between patterns across groups. The use of the 12 distance correlation matrices (dcor) (one for each month) also contributes to what is the main result of the paper, which is to clearly show the different temporal patterns in relation to location and (in some months) altitude. The anomalous behaviors detected at 3 of the 32 stations, namely Valpelline, Fossano, and Aonoch Mòr, can be attributed, respectively, to the facts that Valpelline is the lowest-elevation station in its group; Fossano is the southernmost of the Italian stations, with some sublittoral influence; and Aonoch Mòr has a large number of missing values. In conclusion, these results improve our understanding of temperature spatio-temporal dynamics in two very different regions of Europe and emphasize the importance of consistent analysis of data to assess the ongoing effects of climate change. The intra-annual time patterns of temperature trends could also be compared with climate model results.
- Open Access
- 12 October 2024
- Published by Wiley in International Journal of Climatology
- Vol. 44 (15) , 5398-5417
Analysis of Diurnal Air Temperature Trends and Pattern Similarities in Highland and Lowland Stations of Italy and UK
In this study, an analysis of hourly air temperatures in four groups of 32 stations from the UK highland (5 stations), UK lowland (4 stations), Italian highland (11 stations), and Italian lowland (12 stations) at different altitudes was carried out over the period from 2002 to 2021. The study aimed to examine the trends of each hour of the day during this period, over different averaging time windows (10‐day, 30‐day, and 60‐day). The trends were computed using the Mann–Kendall trend test and Sen's slope estimator. The similarity of trends within and across the groups of stations was assessed using the hierarchical clustering with dynamic time warping technique. An additional analysis was conducted to show the correlation of trends among the group of stations using the correlation distance matrix. Hierarchical clustering and distance correlation analysis show trend similarities and correlations, also indicating dissimilarities among different groups. Using 30‐day averages, significant warming trends in specific months at the Italian stations are evident, especially in February, July, August, and December. The UK highland stations did not show statistically significant trends, but clear pattern similarities were found within the groups, especially in certain months. The ultimate goal of this article is to provide insights into temperature dynamics and climate change characteristics on regional and diurnal scales.
- 12 October 2024
- Published by Wiley in International Journal of Climatology
- Vol. 44 (15) , 5398-5417
Open AccessAnalysis of Diurnal Air Temperature Trends and Pattern Similarities in Highland and Lowland Stations of Italy and UK
In this study, an analysis of hourly air temperatures in four groups of 32 stations from the UK highland (5 stations), UK lowland (4 stations), Italian highland (11 stations), and Italian lowland (12 stations) at different altitudes was carried out over the period from 2002 to 2021. The study aimed to examine the trends of each hour of the day during this period, over different averaging time windows (10‐day, 30‐day, and 60‐day). The trends were computed using the Mann–Kendall trend test and Sen's slope estimator. The similarity of trends within and across the groups of stations was assessed using the hierarchical clustering with dynamic time warping technique. An additional analysis was conducted to show the correlation of trends among the group of stations using the correlation distance matrix. Hierarchical clustering and distance correlation analysis show trend similarities and correlations, also indicating dissimilarities among different groups. Using 30‐day averages, significant warming trends in specific months at the Italian stations are evident, especially in February, July, August, and December. The UK highland stations did not show statistically significant trends, but clear pattern similarities were found within the groups, especially in certain months. The ultimate goal of this article is to provide insights into temperature dynamics and climate change characteristics on regional and diurnal scales.
- Open Access
- 27 March 2023
- Published by Association for Computing Machinery (ACM)
- p. 377-380
Multivariate Time Series Evapotranspiration Forecasting using Machine Learning Techniques
The actual evapotranspiration (AET) could be forecasted using meteorological variables to manage and plan water resources even though it is challenging to choose the relevant variables for prediction. The Pearson correlation method was applied to select candidate variables and further, tolerance and VIF scores are implemented to avoid multicollinearity problems among variables. As a result, five relevant variables are selected for training the AET prediction models. In this paper, we proposed three methods for forecasting AET: (i) deep learning-based (LSTM, GRU, and CNN), (ii) classical machine learning (SVR and RF), and (iii) a statistical technique (SARIMAX). The performance of each model is measured with statistical indicators (RMSE, MSE, MAE, and R2). The results showed that relatively high performance is measured in the LSTM model.
- 27 March 2023
- Published by Association for Computing Machinery (ACM)
- p. 377-380
Open AccessMultivariate Time Series Evapotranspiration Forecasting using Machine Learning Techniques
The actual evapotranspiration (AET) could be forecasted using meteorological variables to manage and plan water resources even though it is challenging to choose the relevant variables for prediction. The Pearson correlation method was applied to select candidate variables and further, tolerance and VIF scores are implemented to avoid multicollinearity problems among variables. As a result, five relevant variables are selected for training the AET prediction models. In this paper, we proposed three methods for forecasting AET: (i) deep learning-based (LSTM, GRU, and CNN), (ii) classical machine learning (SVR and RF), and (iii) a statistical technique (SARIMAX). The performance of each model is measured with statistical indicators (RMSE, MSE, MAE, and R2). The results showed that relatively high performance is measured in the LSTM model.
- Open Access
- 7 December 2021
- Published by Springer Nature in Journal of Big Data
- Vol. 8 (1) , 1-11
Machine learning techniques to predict daily rainfall amount
Predicting the amount of daily rainfall improves agricultural productivity and secures food and water supply to keep citizens healthy. To predict rainfall, several types of research have been conducted using data mining and machine learning techniques of different countries’ environmental datasets. An erratic rainfall distribution in the country affects the agriculture on which the economy of the country depends on. Wise use of rainfall water should be planned and practiced in the country to minimize the problem of the drought and flood occurred in the country. The main objective of this study is to identify the relevant atmospheric features that cause rainfall and predict the intensity of daily rainfall using machine learning techniques. The Pearson correlation technique was used to select relevant environmental variables which were used as an input for the machine learning model. The dataset was collected from the local meteorological office at Bahir Dar City, Ethiopia to measure the performance of three machine learning techniques (Multivariate Linear Regression, Random Forest, and Extreme Gradient Boost). Root mean squared error and Mean absolute Error methods were used to measure the performance of the machine learning model. The result of the study revealed that the Extreme Gradient Boosting machine learning algorithm performed better than others.
- 7 December 2021
- Published by Springer Nature in Journal of Big Data
- Vol. 8 (1) , 1-11
Open AccessMachine learning techniques to predict daily rainfall amount
Predicting the amount of daily rainfall improves agricultural productivity and secures food and water supply to keep citizens healthy. To predict rainfall, several types of research have been conducted using data mining and machine learning techniques of different countries’ environmental datasets. An erratic rainfall distribution in the country affects the agriculture on which the economy of the country depends on. Wise use of rainfall water should be planned and practiced in the country to minimize the problem of the drought and flood occurred in the country. The main objective of this study is to identify the relevant atmospheric features that cause rainfall and predict the intensity of daily rainfall using machine learning techniques. The Pearson correlation technique was used to select relevant environmental variables which were used as an input for the machine learning model. The dataset was collected from the local meteorological office at Bahir Dar City, Ethiopia to measure the performance of three machine learning techniques (Multivariate Linear Regression, Random Forest, and Extreme Gradient Boost). Root mean squared error and Mean absolute Error methods were used to measure the performance of the machine learning model. The result of the study revealed that the Extreme Gradient Boosting machine learning algorithm performed better than others.
- Open Access
- 20 September 2021
- Published by Springer Nature in Research Square
Machine Learning Techniques to Predict Daily Rainfall Amount
It is crucial to predict the amount of daily rainfall to improve agricultural productivities to secure food, and water quality supply to keep the citizen healthy. To predict rainfall, various researches are conducted using data mining and machine learning techniques of different countries’ environmental datasets. The Pearson correlation technique is used to select relevant environmental variables which are used as an input for the machine learning model of this study. The main objective of this study is to identify the relevant atmospheric features that cause rainfall and predict the intensity of daily rainfall using machine learning techniques. The dataset is collected from the local meteorological office to measure the performance of three machine learning techniques as Multivariate Linear Regression, Random Forest and Extreme Gradient Boost. Root mean squared error and Mean absolute Error are used to measure the performance of the machine learning model for this study. The result of the study shows that the Extreme Gradient Boost gradient descent machine learning algorithm performs better than others.
- 20 September 2021
- Published by Springer Nature in Research Square
Open AccessMachine Learning Techniques to Predict Daily Rainfall Amount
It is crucial to predict the amount of daily rainfall to improve agricultural productivities to secure food, and water quality supply to keep the citizen healthy. To predict rainfall, various researches are conducted using data mining and machine learning techniques of different countries’ environmental datasets. The Pearson correlation technique is used to select relevant environmental variables which are used as an input for the machine learning model of this study. The main objective of this study is to identify the relevant atmospheric features that cause rainfall and predict the intensity of daily rainfall using machine learning techniques. The dataset is collected from the local meteorological office to measure the performance of three machine learning techniques as Multivariate Linear Regression, Random Forest and Extreme Gradient Boost. Root mean squared error and Mean absolute Error are used to measure the performance of the machine learning model for this study. The result of the study shows that the Extreme Gradient Boost gradient descent machine learning algorithm performs better than others.