2004Journal of Geophysical Research AtmospheresOpen access

Validation of Multiangle Imaging Spectroradiometer (MISR) aerosol optical thickness measurements using Aerosol Robotic Network (AERONET) observations over the contiguous United States

Yang Liu, Jeremy A. Sarnat, Brent A. Coull, Petros Koutrakis, Daniel J. Jacob

Open full text 106 citations

Abstract

Aerosol optical thickness (AOT) data retrieved by the Multiangle Imaging Spectroradiometer (MISR) in 2001 were compared with AOT measurements from 16 Aerosol Robotic Network (AERONET) sites over the contiguous United States. Overall, MISR and AERONET AOTs were strongly correlated (r = 0.73). Regression analysis showed that the root mean square error (RMSE) of MISR AOT was 0.05. The overall retrieval error of MISR AOT was within ±0.04 ± 0.18 × AOT. This result as well as the regression slope and intercept were comparable with previous results using AOT retrievals from MISR or Moderate Resolution Imaging Spectroradiometer (MODIS). The agreement between MISR and AERONET AOTs was improved (R2 = 0.90, RMSE = 0.04) when data from three western inland sites were excluded. A paired t test indicated that MISR systematically overestimated AOT by 0.02 ± 0.007. It was also shown that this positive bias in MISR AOT was greater during the spring and summer as well as in western United States. Together, these results suggest that MISR AOT measurements may be suitable for quantitative analysis of aerosol abundance. Finally, it is unlikely that the current results will vary when using alternative MISR AOT parameters since our analysis also showed the MISR AOT parameters (best fit, regional mean, and weighted regional mean AOTs) to be highly comparable.

About this research paper

What this paper is about

Aerosol optical thickness (AOT) data retrieved by the Multiangle Imaging Spectroradiometer (MISR) in 2001 were compared with AOT measurements from 16 Aerosol Robotic Network (AERONET) sites over the contiguous United States. Overall, MISR and AERONET AOTs were strongly correlated (r = 0.73). Regression analysis showed that the root mean square error (RMSE) of MISR AOT was 0.05. The overall retrieval error of MISR AOT was within ±0.04 ± 0.18 × AOT. This result as well as the regression slope and intercept were comparable with previous results using AOT retrievals from MISR or Moderate Resolution Imaging Spectroradiometer (MODIS). The agreement between MISR and AERONET AOTs was improved (R2 = 0.90, RMSE = 0.04) when data from three western inland sites were excluded. A paired t test indicated that MISR systematically overestimated AOT by 0.02 ± 0.007. It was also shown that this positive bias in MISR AOT was greater during the spring and summer as well as in western United States. Together, these results suggest that MISR AOT measurements may be suitable for quantitative analysis of aerosol abundance. Finally, it is unlikely that the current results will vary when using alternative MISR AOT parameters since our analysis also showed the MISR AOT parameters (best fit, regional mean, and weighted regional mean AOTs) to be highly comparable.

Why it matters

OpenAlex reports 106 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Aerosol optical thickness (AOT) data retrieved by the Multiangle Imaging Spectroradiometer (MISR) in 2001 were compared with AOT measurements from 16 Aerosol Robotic Network (AERONET) sites over the contiguous United States. Overall, MISR and AERONET AOTs were strongly correlated (r = 0.73). Regression analysis showed that the root mean square error (RMSE) of MISR AOT was 0.05. The overall retrieval error of MISR AOT was within ±0.04 ± 0.18 × AOT. This result as well as the regression slope and intercept were comparable with previous results using AOT retrievals from MISR or Moderate Resolution Imaging Spectroradiometer (MODIS). The agreement between MISR and AERONET AOTs was improved (R2 = 0.90, RMSE = 0.04) when data from three western inland sites were excluded. A paired t test indicated that MISR systematically overestimated AOT by 0.02 ± 0.007. It was also shown that this positive bias in MISR AOT was greater during the spring and summer as well as in western United States. Together, these results suggest that MISR AOT measurements may be suitable for quantitative analysis of aerosol abundance. Finally, it is unlikely that the current results will vary when using alternative MISR AOT parameters since our analysis also showed the MISR AOT parameters (best fit, regional mean, and weighted regional mean AOTs) to be highly comparable.

Key concepts: AERONET, Spectroradiometer, Aerosol, Environmental science, Remote sensing, Linear regression, Mean squared error, Atmospheric sciences

Related papers

Back to paper searchBrowse research topicsOriginal source
Validation of Multiangle Imaging Spectroradiometer (MISR) aerosol optical thickness measurements using Aerosol Robotic Network (AERONET) observations over the contiguous United States — Research Paper | ScholarLens