Scaling of ecosystem phenology across satellite sensors and field-based observations: a case-study of contrasting vegetation types in a semi-arid savanna
| dc.contributor.author | Damant, Richard John | |
| dc.date.accessioned | 2023-11-10T09:56:43Z | |
| dc.date.available | 2023-11-10T09:56:43Z | |
| dc.date.issued | 2022 | |
| dc.description | A dissertation submitted in fulfilment of the requirements for the degree of Master of Science to the Faculty of Science, University of the Witwatersrand, Johannesburg, 2022 | |
| dc.description.abstract | Savannas are characteristically heterogenous and are driven by factors such as rainfall, herbivory, fire, and soil nutrients. All these factors interact differently at different spatial and temporal scales. Savanna phenology is highly unpredictable and is influenced predominantly by rainfall among other environmental factors. Climate change is therefore likely to alter phenological responses which emphasises the need for monitoring. However, the heterogeneous nature of savanna phenology makes it difficult to monitor as savanna treegrass ratios are unpredictable and are not comparable across spatial scales. This similarly applies to the use of remotely sensed data sources used to estimate phenological events, which have different spatial and temporal properties, thus phenological observations are limited to the scales they were derived from. Many studies have made use of the Normalised Difference Vegetation Index (NDVI) to estimate both vegetation biomass and phenological metrics from a variety of different satellite products. However, no studies were found that explicitly investigated how spatial scale influences phenological metric and vegetation biomass estimates in a southern African savanna setting. Therefore, this study aimed to explore the influence of spatial scale on phenology and biomass estimates derived from remotely sensed products in a semi-arid savanna. A spatially nested data collection process was followed from which in situ biomass and remotely sensed NDVI products were collected. Results suggest that finer scale NDVI products are better predictors of herbaceous biomass and, on average, phenological metrics occurred later at coarser scale and earlier at finer spatial scales (~17 days). The conclusions from this study highlight the need for monitoring programs to carefully consider the use of coarse scale NDVI products as a proxy for biomass, particularly in heterogeneous landscapes, and the need to use a multi-scale approach when estimating phenological processes. | |
| dc.description.librarian | PC(2023) | |
| dc.faculty | Faculty of Science | |
| dc.identifier.uri | https://hdl.handle.net/10539/36951 | |
| dc.language.iso | en | |
| dc.school | Animal, Plant and Environmental Sciences | |
| dc.subject | Biomass | |
| dc.subject | Phenology | |
| dc.subject | Savannas | |
| dc.title | Scaling of ecosystem phenology across satellite sensors and field-based observations: a case-study of contrasting vegetation types in a semi-arid savanna | |
| dc.type | Dissertation |