Highlights from the most recent Africa Soil Information Service (AfSIS) reporting period, June-Dec. 2010:

--Signing of the GlobalSoilMap.net consortium agreement by high-level official representatives from each of the participating regional nodes. Most nodes have now begun to either produce first generation map products for large areas, or to at least to identify preferred prediction methods by conducting pilot projects that apply and evaluate alternative mapping methods.

--Several thousand georeferenced soil profile observations, with analytical values for a wide range of
properties, have been added to the ISRIC soil legacy database for Africa (see Fig. 1). Based on these data and spatial covariate layers (e.g., terrain models, climate surfaces, remote sensing data, soil & geological maps ...), work is currently underway to produce the first generation of digital soilmaps to a common global standard. With 30,000 to 40,000 profiles defined as a tangible goal for Africa, additional data holdings for e.g Kenya and Nigeria, are currently being collected and processed, and these will be incorporated in subsequent versions of the maps.

--The ICRAF-ISRIC Soil VNIR (Visible and Near Infrared Reflectance) Spectral Library and associated documentation is downloadable at: http://africasoils.net/data/ICRAF-ISRICspectra. Within the project this valuable reference database is currently being used to test various sampling and pedotransfer nalysis procedures. Updates with e.g. MIR spectra are expected to occur over the coming year.

--Development of online mapping tools for viewing and selecting available AfSIS data has continued. These services utilize open standards for cross-platform data visualizations and were constructed using free and open source software tools. Three custom continent-wide maps for Africa (digital elevation model, wetness index, and specific catchment area) are available at http://africasoils.net for visualization at multiple zoom levels. Downloads of these datasets are also available for the entire continent by quadrant, by country, and by user-defined area. We anticipate that the number of available data layers (e.g. radiometrically corrected Landsat & MODIS satellite images and soil profile measurements) will expand rapidly over the coming 6 months.

--The AfSIS map service was integrated with the GIST_OpenLayers JavaScript library that is under
development at the Center for International Earth Science Information Network (CIESIN) for the map services provided by the NASA data center it hosts, the Socioeconomic Data and Applications Center (SEDAC). This software library simplifies the process of adding new data layers to an OpenLayers map client and features a new layer switcher and improved design.

--Work has continued on strengthening and expanding the IT infrastructure needed to support the AfSIS information system. The AfSIS website, project management and intranet sites have been transferred to a new server that is running on a virtual machine. This virtualization allows the flexibility to run multiple operating systems and allocate additional memory for faster server-side processing as needed. Alpha, beta and production versions of AfSIS web services have been set up for development and both internal and external testing before public release.

--Testing of workflow tools for development, production and delivery of AfSIS data sets, including maps and spectral pedotransfer functions is in progress. The great advantage of these tools is that specialized outputs can be produced on demand, and also that datasets are reproducible and that Figure 1: Legacy soil profile observations with soil carbon measurements (map projection in Lat/Lon, WGS84). Red polygon constitutes AfSIS project area. 300 km Lambert Azimuthal Equal Area grid superimposed for visualizing scale and level of data clustering. the steps to produce them are transparent. Work is in progress to set up scientific workflows to construct digital soil maps for Malawi, Mali and Tanzania. Development of workflow tools is also being supported by development of statistical procedures for soil sampling and mapping using legacy and newly collected data.

--Since February 2010, field surveys of 22 sentinel sites have been completed (see Table 1). The field teams are now functioning well, and the regional scientists and their teams have been instrumental in getting previously delayed survey schedules back on track. At the current sampling pace, we expect to be able to sample 41 sites by month 34, putting us ahead of the originally anticipated schedule.

--All recovered soil samples are being analyzed under common standards in the AfSIS labs in Arusha, Bamako, Lilongwe and Nairobi. The set of standard operating procedures and laboratory manuals describing processes and workflows for all AfSIS lab analyses may be obtained at: http:// afsis.ciesin.columbia.edu/redmine/projects/obj3/documents.

--Regional AfSIS laboratories (Bamako, Lilongwe & Arusha) and at ICRAF headquarters in Nairobi are using a common database for managing sample details and soil information. It is now possible to merge all the regional databases into a full project database. As of 15th November 2010, the database holds information on ~7,000 new soil samples. All of the samples have been scanned using near and mid-infrared spectroscopy, and a subset of reference samples are on track at various stages of analysis. Subsamples of all of the reference samples have also been sent to the USA for microbial DNA analysis.

--Work is underway to migrate all field data collection protocols (e.g. for soil surveys, household surveys and field trials) to an Open Data Kit platform that can be implemented on Android smart phones and tablets (see http://code.google.com/p/opendatakit/). This is expected to decrease data entry efforts substantially and to improve sample tracking and data quality.

--A protocol for implementation of diagnostic trials has been developed for maize, sorghum and paddy rice. There are typically 8 nutrient omission (or addition) treatments implemented in these trials. The protocols have been translated into a comprehensive logbook for field data collection. A protocol for sampling yields obtained under farmers own management is presently under development.

--Results from 121 diagnostic trials that have been conducted in Mali, Malawi and Tanzania have been entered into AfSIS field trial database. These are being analyzed (see Fig. 2). Based on the results obtained, ISFM trials have been set up in Mali and Tanzania.

--A soil fertility management trial database has been set up, and is being populated with legacy data from Kenya, Malawi, Mali, Nigeria and Tanzania.

--Questionnaires were prepared to conduct NAR’s and International end user surveys. The questionaires have been administered via web and national workshop surveys and are currently undergoing analysis.