Can High-Resolution Vegetation Greenness Serve as an Indicator for Stream Drying?
Abstract
Streamflow drying patterns in water-limited regions affect both water quality and quantity, but remain difficult to predict in both time and space. Satellite-derived Normalized Difference Vegetation Index (NDVI) values in riparian evergreen forests may correlate with drying, but these patterns remain unclear in semi-arid riparian deciduous forests. A strong correlation between streamflow and drying may exist if soil moisture is otherwise limited, although there is global isotopic evidence that streams and vegetation derive water from two separate sources, suggesting such a correlation may be weak. Here we test whether riparian deciduous NDVI predicts stream drying using spectral reflectance sensors to measure top-of-canopy NDVI in riparian zones in Murphy Creek, an intermittent, discontinuous stream in semi-arid southern Idaho. We compare ground-based and satellite-based NDVI at varying resolutions to assess the relationship between stream drying and riparian vegetation greenness. Stream drying is assessed spatiotemporally with sensors recording the presence or absence of surface flow along the length of the stream network throughout the seasonal hydrograph recession. Co-located ground-based NDVI measurements centered over each streamflow sensor and on the adjacent banks were compared with 3m, 10 m and 30 m satellite imagery. Results show a weak to moderate correlation between stream dryness and ground-based in-stream NDVI values. We show that ground-based NDVI values vary widely within 1-5m of the stream bank depending on orientation of the sensor and riparian canopy continuity. Thus, any potential relationship between drying and vegetation greenness is weakened by low-resolution satellite imagery that integrates in-stream and near-bank vegetation greenness value. The seasonal NDVI-drying pattern appears to strengthen as the stream network contracts and disconnects, so vegetation greenness that is measured at appropriate resolutions may help predict dynamic surface water availability throughout headwater networks in semi-arid regions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H33H2004B
- Keywords:
-
- 1807 Climate impacts;
- HYDROLOGY;
- 1809 Desertification;
- HYDROLOGY;
- 1813 Eco-hydrology;
- HYDROLOGY;
- 1834 Human impacts;
- HYDROLOGY