RClimTool: A free application for analyzing climatic series
To support decision-making in the agricultural sector many different methodologies and tools have been developed that require climatological information as an input. The use of historical data of sufficient quality and quantity is of vital importance to reduce uncertainty in crop modeling and climate forecasting. It is common to find: typos, missing data, outliers and trends in time series data, which may then require a detailed process of quality control, estimation of missing information and analysis of the series before it is fit for use.
The open access statistical software R, contains packages such as Climdex for quality control and calculation of indicators for signals of climate change and Climatol for the homogenization of climatic series. However, these tools do not integrate all the required analyses to detect and correct anomalies in the series, for this reason the need arises to integrate them into an interface of free access and friendly user. RClimTool has been developed as part of project activities for MADR-CIAT Climate and the Colombian agricultural sector: Adaptation for productive sustainability. It was designed with the aim of facilitating users in statistical analysis for quality control, filling of missing data, homogeneity analysis and calculation of indicators for daily weather series for temperature (maximum and minimum) and precipitation. This is possible through a GUI, developed under the R language, with seven modules (see Figure 1), which offer different options to perform a complete analysis.
Some of the relevant modules are described below:
- Descriptive graphical analysis: Provides a summary of the main features for each series through a descriptive analysis (measures of central tendency and dispersion) and graphics such as box plots, scatter plots and histograms, which allow the visualization of the general behavior of the climatic series.
- Quality control: Some filters to identify unreasonable or incorrect data present in the time series, check internal, temporal and spatial consistency are proposed.
- Filling missing data: The RMAWGEN package is used, which estimates VAR models using information from nearby weather stations to complete the missing data in each station included in the analysis.
- Analysis of homogeneity: Some of the more formal statistical tests are implemented to check the homogeneity of the series, and identify changes in mean and variance. Some of the tests included are Mann Kendall, Mann-Whitney, U Test, F Test, T Test and other formal tests to check normality.
To read more visit the Convenio’s blog post