nerc.ac.uk

The MODIS (collection V006) BRDF/albedo product MCD43D: temporal course evaluated over agricultural landscape

Mira, Maria; Weiss, Marie; Baret, Frédéric; Courault, Dominique; Hagolle, Olivier; Gallego-Elvira, Belen; Olioso, Albert. 2015 The MODIS (collection V006) BRDF/albedo product MCD43D: temporal course evaluated over agricultural landscape. Remote Sensing of Environment, 170. 216-228. 10.1016/j.rse.2015.09.021

Before downloading, please read NORA policies.
[thumbnail of N511940PP.pdf]
Preview
Text
N511940PP.pdf - Accepted Version

Download (1MB) | Preview

Abstract/Summary

The assessment of uncertainties in satellite-derived global surface albedo products is a critical aspect for studying the climate, ecosystem change, hydrology or the Earth's radiant energy budget. However, it is challenged by the spatial scaling errors between satellite and field measurements. This study aims at evaluating the forthcoming MODerate Resolution Imaging Spectroradiometer (MODIS) (Collection V006) Bidirectional Reflectance Distribution Function (BRDF)/albedo product MCD43D over a Mediterranean agricultural area. Here, we present the results from the accuracy assessment of the MODIS blue-sky albedo. The analysis is based on collocated comparisons with higher spatial resolution estimates from Formosat-2 that were first evaluated against local in situ measurements. The inter-sensor comparison is achieved by taking into account the effective point spread function (PSF) for MODIS albedo, modeled as Gaussian functions in the North–South and East–West directions. The equivalent PSF is estimated by correlation analysis between MODIS albedo and Formosat-2 convolved albedo. Results show that it is 1.2 to 2.0 times larger in the East–West direction as compared to the North–South direction. We characterized the equivalent PSF by a full width at half maximum size of 1920 m in East–West, 1200 m in North–South. This provided a very good correlation between the products, showing absolute (relative) Root Mean Square Errors from 0.004 to 0.013 (2% to 7%), and almost no bias. By inspecting 1-km plots homogeneous in land cover type, we found poorer performances over rice and marshes (i.e., relative Root Mean Square Error of about 11% and 7%, and accuracy of 0.011 and − 0.008, respectively), and higher accuracy over dry and irrigated pastures, as well as orchards (i.e., relative uncertainty < 3.8% and accuracy < 0.003). The study demonstrates that neglecting the MODIS PSF when comparing the Formosat-2 albedo against the MODIS one induces an additional uncertainty up to 0.02 (10%) in albedo. The consistency between fine and coarse spatial resolution albedo estimates indicates the ability of the daily MCD43D product to reproduce reasonably well the dynamics of albedo.

Item Type: Publication - Article
Digital Object Identifier (DOI): 10.1016/j.rse.2015.09.021
UKCEH and CEH Sections/Science Areas: Reynard
ISSN: 0034-4257
Additional Keywords: albedo, MODIS, Formosat-2, validation, time series, observation coverage, point spread function, BRDF, narrow-to-broadband, surface reflectance, crop, regional scale
NORA Subject Terms: Agriculture and Soil Science
Meteorology and Climatology
Date made live: 21 Oct 2015 13:03 +0 (UTC)
URI: https://nora.nerc.ac.uk/id/eprint/511940

Actions (login required)

View Item View Item

Document Downloads

Downloads for past 30 days

Downloads per month over past year

More statistics for this item...