On 30 April 2017, the funding period of the IDESSA project has officially come to an end. Please refer to the publications section for the final reports which are now available. The project resulted in major advancements regarding the development of a tool supporting decision making in sustainable savanna management.
The software underlying the decision-support system - the VAT system (Visualization, Analysis and Transformation) - is available as open source and can currently be downloaded from this homepage in the results section. In the future, the project will be moved to GitHub (https://github.com/) in order to further develop the software and to involve computer scientists from other working groups in the development of the VAT system, thus ensuring the sustainability of the project. At present, all steps in the development of the monitoring products can be accessed as R-scripts via https://github.com/environmentalinformatics-marburg/magic/tree/master/IDESSA. The most recent version of overarching System and data sets can be found at http://idessa.org/?page=demo. All data and functions will be made available on a server at the University of Marburg (dbsvm.mathematik.uni-marburg.de:32768/) and should be supplemented by additional records. The developments will be further used by other projects in the frame of the German Federation for Biological Data GFBio (www.gfbio.org) and the Group on Earth Observations GEOBON (http://geobon.org/).
The IDESSA team thanks the German Ministry of Education and Research (BMBF) for funding and all collaborators. We are very grateful for all the support we've received from involved stakeholders and the project executing organization. Most of the conducted field work, the parameterization and validation of developed models, and steps towards the finalization and implementation of IDESSA heavily relied on South African scientific expert inputs and the practical knowledge and long-term experience of local land users and reserve managers. Without the trusting cooperation with this large and diverse group of scientists and stakeholders from the civil society, the successful implementation of the project would not have been possible.
We are keen to further develop single components of the decision-support system in future research and look forward to publishing more outcomes of this IDESSA project. If you are interested in potential collaborations or have questions relating to certain research aspects, please don't hesitate to contact us.
IDESSA held a 1-day workshop in the Molopo- (Ganyesa, North-West Province) and Mier (Askham, Northern Cape Province) region, both core study areas of IDESSA within the southwestern Kalahari.
Similar to the 2016 scientific feedback workshop, the agenda was to discuss the functionality of the IDESSA decision-support system (DSS) with potential end users, this time giving special attention to user-friendliness and acceptability by practitioners. Accordingly, the workshops were performed to specifically get through to the diverse group of farmers, governmental experts (agricultural sector) and land managers. We received very valuable inputs and suggestions for improvement, while the development of the DSS was clearly acknowledged.
Hourly 3 km resolution satellite-based rainfall estimates are now available for Southern Africa for the years 2010-2014.
The rainfall retrieval developed in IDESSA subproject 1 is based on data from Meteosat Second Generation and a neural network approach. The model was calibrated and validated using rainfall measurements from about 350 weather stations from SAWS, SASSCAL and IDESSA of the years 2010-2014. The validation indicates that the model is very well able to detect rainy clouds. Predicted hourly rainfall quantities could be estimated with an average hourly correlation of rho = 0.33 and a RMSE of 0.72. The correlation increases with temporal aggregation to 0.52 (daily), 0.67 (weekly) and 0.71 (monthly). We compared the results to the IMERG product of the Global Precipitation Measurement (GPM) mission. The new retrieval outperformed GPM IMERG in terms of rainfall area detection, where GPM IMERG considerably underestimated rainfall events and had a similar performance in terms of rainfall quantities but in a higher spatial resolution. If being interested, please contact IDESSA subproject 1 for hourly rainfall data (or daily/monthly aggregates), currently for the years 2010-2014. The data cover South Africa, Lesotho and Swaziland, Namibia, Botswana, Zimbabwe, as well as parts of Mozambique. A publication that describes the methods and results in more detail is currently in progress. The figure below shows the estimated monthly precipitation sums for the year 2013.
Dr. Theunis Morgenthal, scientist at the South African Department of Agriculture, Forestry and Fisheries in Pretoria, South Africa, visited the universities Marburg and Göttingen for short-term research within IDESSA thanks to a granted DAAD scholarship. Preliminary results of the collaborative research was presented by him at the 46th annual meeting of the Ecological Society of Germany, Austria and Switzerland in Marburg (see publications).
IDESSA held a scientific feedback workshop in Kimberley with 15 invited participants from South Africa. Thanks to the different backgrounds, knowledge and expertise, we received a critical and multifaceted evaluation of the three components of the IDESSA decision-support system. Discussions and group work were very stimulating, and related output will be used to improve our approach and adapt developed models.
Thomas Higginbottom, PhD student of Elias Symeonakis from Manchester Metropolitan University, UK, working in the LanDDApp project (Land Degradation and Desertification Appraisal for South Africa) visited the Team of IDESSA subproject 1 in Marburg. The purpose was to jointly work on combining aerial images, Google images, Landsat and MODIS data to monitor bush encroachment in South Africa. We are looking forward to extend our cooperation and to compare the capabilities of Landsat and MODIS-based time series for our tasks.
A total of 21 scientists and practitioners from different universities and organizations met at the Potchefstroom campus of the North-West University to discuss problems related to bush thickening of savannas in South Africa. A focus was on remote sensing- and ground-based approaches for the assessment and monitoring of these vegetation dynamics.
The workshop was followed by a two-day field trip to some sites in the Molopo savanna of the North-West province, where field work is carried out by scientists of IDESSA and READ (Department of Rural, Environmental and Agricultural Development).
German representatives of IDESSA visited the South African collaboration partners and went on a joint field trip. During their stay, they travelled the core study areas in three provinces, interviewed land users and other stakeholders and set up 15 weather stations.