Accuracy and Uncertainty in OLCI Terrestrial Chlorophyll Index (OTCI): Lessons Learn from MTCI-EVAL Project
Dash, Jadunandan1; Vuolo, Francesco2; Watmough, Gary3; Frampton, Will1; Curran, Paul4; Lajas, Dulce5; Kwiatkowska, Ewa6
1University of Southampton, UNITED KINGDOM; 2University of Natural Resources and Life Sciences (BOKU), Peter Jordan Str. 82, A-1190 Vienna, AUSTRIA; 3Earth Institute,Columbia University, UNITED STATES; 4City University, London, UNITED KINGDOM; 5ESTEC,ESA, NETHERLANDS; 6EUMETSAT, GERMANY
A measure of canopy chlorophyll content was operationally available at global scale to users through the MERIS Terrestrial Chlorophyll Index or MTCI product from ENVISAT MERIS. MTCI has shown potential to be used in several applications ranging from mapping the effect of salt stress in coastal vegetation, mapping vegetation phenology to estimate gross primary productivity for different biomes. ESA's Sentinel-3 OTCI will guarantee the continuity of MTCI at the MERIS spatial resolution for global scale studies.
In the MTCI-EVAL project data from model, greenhouse experiment and dedicated filed campaign were used to identify uncertainties in the product due to various sensor and non-sensor variables. While estimating uncertainty in the product we adopted the process suggested by the Quality Assurance Framework for Earth Observation (QA4EO). This in turn was useful in defining specific binary quality flags for the OTCI product and can be easily implemented in the processing chain. In preparation for launch of the S3 mission, the performance of MTCI needs to be understood and its effectiveness to provide robust measures of canopy chlorophyll content. Validation of this vegetation index is also important across a range of vegetation types, spatial scales and different environmental conditions. Therefore, in total data from 6 validation campaigns (4 from previous ESA campaign and 2 dedicated campaigns from MTCI EVAL project) were used to find relationship between MTCI and ground chlorophyll content. Despite the difference in vegetation types, density, growth stages and environmental condition, there was a strong positive relationship between MTCI and ground chlorophyll content. The project resulted in a detail protocol document for validation moderate resolution biophysical products which can be used for future sentinel 3 validation activities.