Abstract
In developing countries like Ethiopia where there is abundant water resources potential and also luck of reliable meteorological quality data, it expected to face the problem of missing meteorological data. Therefore, in conducting any water resources studies in any river basin for water resource project planning and management (like small scale irrigation), the first step before starting data analysis is to fill up the missing values of the meteorological variables (like rainfall, temperature, sunshine, wind speed etc.) which are required to start the study. One way of filling these missing variables is using datasets from other stations in the surrounding and applying appropriate spatial interpolation methods. A lot of studies have been conducted around the world to identify which method is the best to be applied to particular study area among the available spatial interpolation techniques. But when we come to Ethiopia, the study area, few or no studies are conducted to recommend the best performed method. Therefore, the objective of this paper is to conduct comparative evaluation of five interpolation techniques Nearest Neighbour (NN), Inverse Distance Weighting Average (IDWA), Modified Inverse Distance Weighting Average (MIDWA), Kriging Method (KM) and Thin Plate Spline (TPS) for estimation of four climatic variables (rainfall, mean temperature, wind speed and sunshine fraction) over complex topography of Ethiopia. Performance assessment is done using Mean Error (ME), Mean Absolute Error (MAE), Mean Relative Error (MRE) and Root Mean Square Error (RMSE); and the number of the meteorological stations selected for validation is ten (10) and these are distributed over the study area taking into account the variation of elevation ranging from 860 m (Awash) to 2420 m (Debremarkos) above sea level. The radial distances of 100 km and 200 km were selected and it was found that 100 km radial distance was not appropriate to compare all methods as some variables could not be estimated by KM and TPS. Therefore, 200 km was selected for further analysis and the result showed that NN, IDWA, and MIDWA were best methods relative to the remaining two methods (KM and TPS) for all variables and all stations except at Dire Dawa and Addis Ababa-Bole for estimation of wind speed using all methods except NN, and rainfall using TPS, respectively. Hence, NN, IDWA, and MIDWA methods could be used for estimation of missing meteorological variables over Ethiopia whenever necessary.
Highlights
In developing countries like Ethiopia where there is sufficient amount of water resources potential and luck of high quality meteorological data, it expected to have the problem of facing missing meteorological data even though all means were used to avoid these missing values from the records [1] [2]
Performance assessment is done using Mean Error (ME), Mean Absolute Error (MAE), Mean Relative Error (MRE) and Root Mean Square Error (RMSE); and the number of the meteorological stations selected for validation is ten (10) and these are distributed over the study area taking into account the variation of elevation ranging from 860 m (Awash) to 2420 m (Debremarkos) above sea level
Radius of Influence, r = 100 km According to recommendation of [23], as the first attempt, the radius of influence was specified as 100 km and some observation was made and the following results were obtained: First, for common radial distance, r = 100 km,the number of sampling stations selected for all variables, for all stations, and for all interpolation methods is the same and it is 10
Summary
In developing countries like Ethiopia where there is sufficient amount of water resources potential and luck of high quality meteorological data, it expected to have the problem of facing missing meteorological data even though all means were used to avoid these missing values from the records [1] [2]. During conducting any hydrological and environmental studies/modeling in any river basin for water resources project planning and management, the first step before starting data analysis is to fill up the missing values of the meteorological variables (like rainfall, temperature, sunshine, wind speed etc.) required for the study [5]. Unless these missing values are filled up by appropriate methods so as to make the dataset complete, the data used may be biased and can lead to wrong conclusions [6]. In order to get good results of estimation, it is a must to select best interpolation technique for the study selected as the interpolation techniques are affected by different factors like sample size of the data, the sampling design and data properties [8]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
