Analyzing the feasibility of a space-borne sensor (SPOT-6) to estimate the height of submerged aquatic vegetation (SAV) in inland waters

Abstract

Remote sensing based approaches have been widely used over the years to monitor and manage submerged aquatic vegetation (SAV) or aquatic macrophytes mainly by mapping their spatial distribution and at the most, modeling SAV biomass. Remote sensing based studies to map SAV heights are rare because of the complexities in modeling water column optical proprieties. SAV height is a proxy for biomass and can be used to estimate plant volume when combined with percent cover. The objective of this study was to explore the feasibility of a satellite sensor to estimate the SAV height distribution in an inland reservoir. Also to test different radiative transfer theory based bio-optical models for estimating SAV heights using SPOT-6 data. The satellite-based multispectral data have rarely been used and SPOT-6 data, to the best of our knowledge, have never been used to estimate SAV heights in inland water bodies. In addition to depth and hydroacoustic data, in situ hyperspectral radiance and irradiance were measured at different depths to compute remote sensing reflectance (R rs) and the attenuation coefficients (K d and K Lu). Two models, Palandro et al. (2008) and Dierssen et al. (2003), were used to derive bottom reflectance from both in situ and atmospherically corrected SPOT-6 R rs. Bottom reflectance-based vegetation indices (green-red index, slope index, and simple ratio) were used to estimate SAV heights. Validation was performed using echosounder acquired hydroacoustic data. In situ model calibration produced an R 2 of 0.7, however, the validation showed a systematic underestimation of SAV heights and high Root Mean Square Error (RMSE); indicating that there is a greater sensitivity in in situ models to localized variations in water column optical properties. The model based on SPOT-6 data presented higher accuracy, with R 2 of 0.54 and RMSE of 0.29 m (NRMSE = 15%). Although the models showed a decreased sensitivity for SAVs at depths greater than 5 m with a height of 1.5 m, the finding nonetheless is significant because it proves that re-calibration of existing bottom reflectance models with more field data can enhance the accuracy to be able to periodically map SAV heights and biomass in inland waters. Although the initial results presented in this study are encouraging, further calibration of the model is required across different species, seasons, sites, and turbidity regime in order to test its application potential.

Remote sensing of submerged aquatic vegetation: an introduction and best practices review

“…Remote sensing has been applied to determine SAV distribution, cover classes, canopy density, health, and species. Assessing the biophysical properties of SAV through RS has yet to be extensively explored, and those studies that did, examined superficial characteristics such as biomass or plant height (Malthus 2017;Rotta et al 2018). Measurements from aquatic RS have also been used as inputs for modelling fish habitat distributions (Haghi Vayghan et al 2013).…”

“…Sensors record reflected radiance from the surface (and in-scattered from nearby objects) in the form of digital numbers (DN), with the range of possible DN values corresponding to the radiometric resolution. Radiometric correction converts the raw DN data to radiance for the user, which is the amount of energy reaching the sensor, given in Spectral Radiance Units (1 SRU = 1 µWcm -1 sr -1 nm -1 ) (Ripley et al 2009). This correction accounts for the sensor-specific detection and sensitivity variations and is often done using calibration files provided by the sensor manufacturer or another calibration provider (Lekki et al 2017).…”

Kd(PAR) and a Depth Based Model to Estimate the Height of Submerged Aquatic Vegetation in an Oligotrophic Reservoir: A case study at Nova Avanhandava

“…In situ Rrs spectra acquired by RAMSES/TriOS sensors were predicted to match the SPOT-6 sensor bands and were compared with SPOT-6 atmospherically corrected sampling points for performance validation. More details of atmospheric correction of SPOT-6 in Rotta et al [16] and Rotta et al [26]. The spatial distribution of Kd(PAR) was divided into 12 classes based on the minimum and maximum values.…”

Multi-Source Evidence Data Fusion Approach to Detect Daily Distribution and Coverage of Ulva Prolifera in the Yellow Sea, China

“…MODIS (1-day repeat cycle) has a notably higher observation frequency than Landsat (16-day repeat cycle) [15]. The Sentinel series of satellites, with a 10-60 m resolution and a 10-day visit cycle [16], [17], the WorldView series of satellites with a 3.5-meter resolution and a 1.7-day visit cycle on average [16], [18], and the SPOT-6/7 satellites with a 10-meter resolution and a 26-day visit cycle [1], [19] have higher U. prolifera detection accuracies than MODIS under normal circumstances. Among these RS data, the MODIS and Sentinel series of satellites are open, free, and easy to acquire and with suitable visit cycle in U. prolifera detection.…”