High resolution satellite and reanalysis-based air temperature estimates have huge potential to complement the sparse networks of air temperature measurements from ground stations in Africa. The recently released Climate Hazards Center Infrared Temperature with Stations (CHIRTS-daily) dataset provides daily minimum and maximum air temperature estimates on a near-global scale from 1983 to 2016. This study assesses the performance of CHIRTS-daily in comparison with measurements from eight ground stations in diverse locations across Africa from 1983 to 2016, benchmarked against the ERA5 and ERA5-Land reanalysis to understand its potential to provide localized temperature information. Compared to ERA5 and ERA5-Land, CHIRTS-daily maximum temperature has higher correlation and lower bias of daily, annual mean maximum and annual extreme maximum temperature. It also exhibits significant trends in annual mean maximum temperature, comparable to those from the station data. CHIRTS-daily minimum temperatures generally have higher correlation, but larger bias than ERA5 and ERA5-Land. However, the results indicate that CHIRTS-daily minimum temperature biases may be largely systematic and could potentially be corrected for. Overall, CHIRTS-daily is highly promising as it outperforms ERA5 and ERA5-Land in many areas, and exhibits good results across a small, but diverse set of sites in Africa. Further studies in specific geographic areas could help support these findings.

The African Data Initiative started as a crowd-sourced campaign to improve the teaching of statistics in African universities. The analysis of climate data provides one suitable context to illustrate ideas that lead to a radical new form of teaching. The problem within the context comes first, the technicalities are largely reduced – mathematics is supported by meta knowledge and backed up by modelling; calculations are reduced by user-friendly software that is also used by experts. The problems are treated similarly to research questions and the results are often easier to interpret, making sense as potential answers in their context. The criteria of this approach are compared to the framework proposed by W. G. Cobb to reform statistics education in the light of the latest developments in statistics, driven by the huge increase of data. Implementation details are presented around three components: case studies, data, and the required skills. Together, these three components describe an alternative education pathway centred around statistical problem solving. The focus on interpretations of results within a real context enables software, mathematical thinking and modelling to play a supportive role, which flattens the prerequisites of complex methods and encourages their use across all levels of education. First published February 2020 at Statistics Education Research Journal Archives

High‐resolution rainfall estimates from satellite and reanalysis data sources (SRE) could play a major role in improving climate services for agriculture. This is particularly relevant in regions that predominantly rely on rain‐fed farming but lack a dense network of ground‐based measurements to provide localised historical climate information, as in most of the Global South. However, there is a need for a framework which practitioners can use to determine the suitability of these estimated data for specific agricultural applications. This paper presents a comprehensive methodology for evaluating the ability of SRE to provide historical rainfall information for agricultural applications, primarily through comparison with ground‐based measurements. The methodology comprises five main steps: data selection and pre‐processing, spatial and temporal consistency checks, quantitative SRE‐gauge comparisons, bias correction, and application specific summaries. The methodology makes use of graphical summaries, standard comparison metrics, and Markov chain models. We describe how users can apply this methodology to evaluate rainfall estimates for specific applications, complementing existing validation studies. Evaluation cases are presented to demonstrate the methodology using five widely used satellite and reanalysis rainfall products and ground‐based measurements from 12 stations in Africa and the Caribbean. The case studies demonstrate how the methodology can be applied to examine multiple aspects of the rainfall estimates. While previous validation studies ask “Does the SRE estimate the true rainfall well?”, this methodology provides means of establishing “To what extent can an SRE be used for this specific purpose?” and a comprehensive framework for achieving this. This meets a major need for location specific rainfall information to improve climate information services for millions of small‐holder farming households.

High resolution satellite and reanalysis-based air temperature estimates have huge potential to complement the sparse networks of air temperature measurements from ground stations in Africa, and support vital studies in health, agriculture, diseases, and climate change. The recently released Climate Hazards Center Infrared Temperature with Stations (CHIRTS-daily) dataset provides daily minimum and maximum air temperature estimates on a near-global scale from 1983 to 2016. This study assesses the performance of CHIRTS-daily in comparison with measurements from eight ground stations in diverse locations across Africa from 1983 to 2016 and benchmarked against the ERA5 and ERA5-Land reanalysis in order to understand its potential to provide localised temperature information. CHIRTS-daily maximum temperature exhibits good performance across the sites, measures and time scales. Compared to ERA5 and ERA5-Land, CHIRTS-daily has higher correlation and lower bias of daily, annual mean maximum and annual extreme maximum temperature, often within 1.0° on average. It also exhibits significant trends in annual mean maximum temperature, comparable to those from the station data. CHIRTS-daily minimum temperatures generally has higher correlation, but larger bias than ERA5 and ERA5-Land, at more than 2° at four sites. However, the results indicate that CHIRTS-daily minimum temperature biases may be largely systematic, and therefore could potentially be corrected for. Overall, CHIRTS-daily is highly promising as it outperforms ERA5 and ERA5-Land in many areas, and exhibits good results across a small, but diverse set of sites in Africa. Further studies to understand performance in specific geographic areas and terrains could help support these findings.

Each year increasing amounts of data are being produced and there are growing trends towards data becoming more accessible, particularly online. Here we present a range of examples where data are conveniently arranged in multiple linked rectangles or data frames. They are often omitted from all but advanced statistics courses. However, they are common in practice, hence their omission leaves graduates poorly prepared for real world problems. The obvious example is a survey that is at multiple levels. Other examples include multiple time series with spatial data, where the spatial information is in a separate data frame; and data sets in a single rectangle (data frame) but where the analyses are on summary data. The statistical software, R-Instat, resulting from the African Data Initiative is designed to make it easy to handle such data.

This poster focusses on the describe menu of R-Instat; in particular, features relating to data exploration. The exploration of data is fundamentally important for anyone working with data, yet it receives relatively little importance in many statistics courses. The poster illustrates some key aspects of tables and graphs in R-Instat and how they can be used to quickly get an overview of data. It shows how descriptive analyses are made into an easy and natural process, both to teach and to learn. We illustrate a wide variety of graphs that use ggplot2 in R, how to produce tables that are eye-catching and useful, and the similarities between tables and graphs. We can have a table of graphs or graphs where some of the cells have a table. Particular attention will be drawn to the value of a true problem based approach, where students start with a problem and some data and are free to use any appropriate analytic methods. The key to this approach is to have data which is rich enough that a combination of approaches yields more insights than any single analysis. Analyzing data can be fun!