Consistency of vegetation index seasonality across the Amazon rainforest

Eduardo Eiji Maeda (Corresponding Author), Yhasmin Mendes Moura, Fabien Wagner, Thomas Hilker, Alexei I. Lyapustin, Yujie Wang, Jérôme Chave, Matti Mõttus, Luiz E.O.C. Aragão, Yosio Shimabukuro

Research output: Contribution to journalArticleScientificpeer-review

13 Citations (Scopus)

Abstract

Vegetation indices (VIs) calculated from remotely sensed reflectance are widely used tools for characterizing the extent and status of vegetated areas. Recently, however, their capability to monitor the Amazon forest phenology has been intensely scrutinized. In this study, we analyze the consistency of VIs seasonal patterns obtained from two MODIS products: the Collection 5 BRDF product (MCD43) and the Multi-Angle Implementation of Atmospheric Correction algorithm (MAIAC). The spatio-temporal patterns of the VIs were also compared with field measured leaf litterfall, gross ecosystem productivity and active microwave data. Our results show that significant seasonal patterns are observed in all VIs after the removal of view-illumination effects and cloud contamination. However, we demonstrate inconsistencies in the characteristics of seasonal patterns between different VIs and MODIS products. We demonstrate that differences in the original reflectance band values form a major source of discrepancy between MODIS VI products. The MAIAC atmospheric correction algorithm significantly reduces noise signals in the red and blue bands. Another important source of discrepancy is caused by differences in the availability of clear-sky data, as the MAIAC product allows increased availability of valid pixels in the equatorial Amazon. Finally, differences in VIs seasonal patterns were also caused by MODIS collection 5 calibration degradation. The correlation of remote sensing and field data also varied spatially, leading to different temporal offsets between VIs, active microwave and field measured data. We conclude that recent improvements in the MAIAC product have led to changes in the characteristics of spatio-temporal patterns of VIs seasonality across the Amazon forest, when compared to the MCD43 product. Nevertheless, despite improved quality and reduced uncertainties in the MAIAC product, a robust biophysical interpretation of VIs seasonality is still missing.
Original languageEnglish
Pages (from-to)42-53
JournalInternational Journal of Applied Earth Observation and Geoinformation
Volume52
DOIs
Publication statusPublished - 2016
MoE publication typeA1 Journal article-refereed

Fingerprint

vegetation index
rainforest
seasonality
atmospheric correction
MODIS
reflectance
Microwaves
Availability
product
litterfall
clear sky
Ecosystems
phenology
Remote sensing
pixel
Contamination
Lighting
Productivity
Pixels
Calibration

Cite this

Maeda, E. E., Moura, Y. M., Wagner, F., Hilker, T., Lyapustin, A. I., Wang, Y., ... Shimabukuro, Y. (2016). Consistency of vegetation index seasonality across the Amazon rainforest. International Journal of Applied Earth Observation and Geoinformation, 52, 42-53. https://doi.org/10.1016/j.jag.2016.05.005
Maeda, Eduardo Eiji ; Moura, Yhasmin Mendes ; Wagner, Fabien ; Hilker, Thomas ; Lyapustin, Alexei I. ; Wang, Yujie ; Chave, Jérôme ; Mõttus, Matti ; Aragão, Luiz E.O.C. ; Shimabukuro, Yosio. / Consistency of vegetation index seasonality across the Amazon rainforest. In: International Journal of Applied Earth Observation and Geoinformation. 2016 ; Vol. 52. pp. 42-53.
@article{e5a1890a5db145f78a8f164dbfbf013b,
title = "Consistency of vegetation index seasonality across the Amazon rainforest",
abstract = "Vegetation indices (VIs) calculated from remotely sensed reflectance are widely used tools for characterizing the extent and status of vegetated areas. Recently, however, their capability to monitor the Amazon forest phenology has been intensely scrutinized. In this study, we analyze the consistency of VIs seasonal patterns obtained from two MODIS products: the Collection 5 BRDF product (MCD43) and the Multi-Angle Implementation of Atmospheric Correction algorithm (MAIAC). The spatio-temporal patterns of the VIs were also compared with field measured leaf litterfall, gross ecosystem productivity and active microwave data. Our results show that significant seasonal patterns are observed in all VIs after the removal of view-illumination effects and cloud contamination. However, we demonstrate inconsistencies in the characteristics of seasonal patterns between different VIs and MODIS products. We demonstrate that differences in the original reflectance band values form a major source of discrepancy between MODIS VI products. The MAIAC atmospheric correction algorithm significantly reduces noise signals in the red and blue bands. Another important source of discrepancy is caused by differences in the availability of clear-sky data, as the MAIAC product allows increased availability of valid pixels in the equatorial Amazon. Finally, differences in VIs seasonal patterns were also caused by MODIS collection 5 calibration degradation. The correlation of remote sensing and field data also varied spatially, leading to different temporal offsets between VIs, active microwave and field measured data. We conclude that recent improvements in the MAIAC product have led to changes in the characteristics of spatio-temporal patterns of VIs seasonality across the Amazon forest, when compared to the MCD43 product. Nevertheless, despite improved quality and reduced uncertainties in the MAIAC product, a robust biophysical interpretation of VIs seasonality is still missing.",
author = "Maeda, {Eduardo Eiji} and Moura, {Yhasmin Mendes} and Fabien Wagner and Thomas Hilker and Lyapustin, {Alexei I.} and Yujie Wang and J{\'e}r{\^o}me Chave and Matti M{\~o}ttus and Arag{\~a}o, {Luiz E.O.C.} and Yosio Shimabukuro",
year = "2016",
doi = "10.1016/j.jag.2016.05.005",
language = "English",
volume = "52",
pages = "42--53",
journal = "International Journal of Applied Earth Observation and Geoinformation",
issn = "1569-8432",
publisher = "Elsevier",

}

Maeda, EE, Moura, YM, Wagner, F, Hilker, T, Lyapustin, AI, Wang, Y, Chave, J, Mõttus, M, Aragão, LEOC & Shimabukuro, Y 2016, 'Consistency of vegetation index seasonality across the Amazon rainforest', International Journal of Applied Earth Observation and Geoinformation, vol. 52, pp. 42-53. https://doi.org/10.1016/j.jag.2016.05.005

Consistency of vegetation index seasonality across the Amazon rainforest. / Maeda, Eduardo Eiji (Corresponding Author); Moura, Yhasmin Mendes; Wagner, Fabien; Hilker, Thomas; Lyapustin, Alexei I.; Wang, Yujie; Chave, Jérôme; Mõttus, Matti; Aragão, Luiz E.O.C.; Shimabukuro, Yosio.

In: International Journal of Applied Earth Observation and Geoinformation, Vol. 52, 2016, p. 42-53.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Consistency of vegetation index seasonality across the Amazon rainforest

AU - Maeda, Eduardo Eiji

AU - Moura, Yhasmin Mendes

AU - Wagner, Fabien

AU - Hilker, Thomas

AU - Lyapustin, Alexei I.

AU - Wang, Yujie

AU - Chave, Jérôme

AU - Mõttus, Matti

AU - Aragão, Luiz E.O.C.

AU - Shimabukuro, Yosio

PY - 2016

Y1 - 2016

N2 - Vegetation indices (VIs) calculated from remotely sensed reflectance are widely used tools for characterizing the extent and status of vegetated areas. Recently, however, their capability to monitor the Amazon forest phenology has been intensely scrutinized. In this study, we analyze the consistency of VIs seasonal patterns obtained from two MODIS products: the Collection 5 BRDF product (MCD43) and the Multi-Angle Implementation of Atmospheric Correction algorithm (MAIAC). The spatio-temporal patterns of the VIs were also compared with field measured leaf litterfall, gross ecosystem productivity and active microwave data. Our results show that significant seasonal patterns are observed in all VIs after the removal of view-illumination effects and cloud contamination. However, we demonstrate inconsistencies in the characteristics of seasonal patterns between different VIs and MODIS products. We demonstrate that differences in the original reflectance band values form a major source of discrepancy between MODIS VI products. The MAIAC atmospheric correction algorithm significantly reduces noise signals in the red and blue bands. Another important source of discrepancy is caused by differences in the availability of clear-sky data, as the MAIAC product allows increased availability of valid pixels in the equatorial Amazon. Finally, differences in VIs seasonal patterns were also caused by MODIS collection 5 calibration degradation. The correlation of remote sensing and field data also varied spatially, leading to different temporal offsets between VIs, active microwave and field measured data. We conclude that recent improvements in the MAIAC product have led to changes in the characteristics of spatio-temporal patterns of VIs seasonality across the Amazon forest, when compared to the MCD43 product. Nevertheless, despite improved quality and reduced uncertainties in the MAIAC product, a robust biophysical interpretation of VIs seasonality is still missing.

AB - Vegetation indices (VIs) calculated from remotely sensed reflectance are widely used tools for characterizing the extent and status of vegetated areas. Recently, however, their capability to monitor the Amazon forest phenology has been intensely scrutinized. In this study, we analyze the consistency of VIs seasonal patterns obtained from two MODIS products: the Collection 5 BRDF product (MCD43) and the Multi-Angle Implementation of Atmospheric Correction algorithm (MAIAC). The spatio-temporal patterns of the VIs were also compared with field measured leaf litterfall, gross ecosystem productivity and active microwave data. Our results show that significant seasonal patterns are observed in all VIs after the removal of view-illumination effects and cloud contamination. However, we demonstrate inconsistencies in the characteristics of seasonal patterns between different VIs and MODIS products. We demonstrate that differences in the original reflectance band values form a major source of discrepancy between MODIS VI products. The MAIAC atmospheric correction algorithm significantly reduces noise signals in the red and blue bands. Another important source of discrepancy is caused by differences in the availability of clear-sky data, as the MAIAC product allows increased availability of valid pixels in the equatorial Amazon. Finally, differences in VIs seasonal patterns were also caused by MODIS collection 5 calibration degradation. The correlation of remote sensing and field data also varied spatially, leading to different temporal offsets between VIs, active microwave and field measured data. We conclude that recent improvements in the MAIAC product have led to changes in the characteristics of spatio-temporal patterns of VIs seasonality across the Amazon forest, when compared to the MCD43 product. Nevertheless, despite improved quality and reduced uncertainties in the MAIAC product, a robust biophysical interpretation of VIs seasonality is still missing.

UR - http://www.scopus.com/inward/record.url?eid=2-s2.0-84997795123&partnerID=MN8TOARS

U2 - 10.1016/j.jag.2016.05.005

DO - 10.1016/j.jag.2016.05.005

M3 - Article

VL - 52

SP - 42

EP - 53

JO - International Journal of Applied Earth Observation and Geoinformation

JF - International Journal of Applied Earth Observation and Geoinformation

SN - 1569-8432

ER -