Triple Sensor Collocation can be used to validate 3 independent observations at a location, when the error-free true value is not known. With this you can judge, which water or climate observation, i.e. your citizen observation, conventional station data or a remotely sensed satellite look-up is most reliable.
Stockholm Demo Use case: rainfall at the Dano area
The Afrialliance web-demo illustrates the technique for rainfall observations (July 2015) near Dano in Southwest Burkina Faso, a research area of West African Afrialliance partner WASCAL. Citizen locations were extracted from the Open access WPDx or Water Point Database. Citizen names were adapted (privacy) and rain data were generated for demo purposes. Observed meteorological station data were obtained from WASCAL.org and CHIRPS2 was used as open access satellite precipitation. This AA demonstration package is under development and will also be available on-line via the AA project portal www.afrialliance.orgRun the Demo
Change the observation period and view which data source is most reliable in that period (compare the color of the symbols to the color triangle legend). Scroll map and move mouse pointer over Citizen location(s) for evaluation statistics.Method and references
The method is based on statistical covariance analysis on three independent data sources and yields the errors (RMSE) and correlations and permits to rank the dataset reliability.McColl, K. & Vogelzang, J., Konings, A., Entekhabi, D., Piles, M. & Stoffelen, A. (2014). Extended Triple Collocation: estimating errors and correlation coefficients with respect to an unknown target. Geophysical Research Letters. 10.1002/2014GL061322. [link]
Mannaerts, C.M., Maathuis, B., Wehn, U., Gerrets, T., Riedstra, H., Becht, R. and Lemmens, R. (2017) - Afrialliance Project Report Deliverable 4.4: Constraints and Opportunities for Water Resources Monitoring & Forecasting using the Triple Sensor approach. Available here [link]
Data sources: Wascal (2018) http://www.wascal.org ; Chirps (2018) http://chg.geog.ucsb.edu/data/ ; WPDx (2018) http://www.waterpointdata.org
About: Demo-app developed for Afrialliance by consortium partner ITC (Faculty of Geo-information Sciences & Earth Observation, Department of Water Resources, UTwente). For more information, pls. contact C. Mannaerts c.m.m.mannaerts@utwente.nl, V. Retsios v.retsios@utwente.nl and R. Lemmens r.l.g.lemmens@utwente.nl .
The Afrialliance project has received funding from the EU H2020 program under grant agreement No. 689162. For more project info, pls. contact Uta Wehn u.wehn@un-ihe.org .
