Emissivity and reflection model for calculating unpolarized isotropic water surface-leaving radiance in the infrared. 2: Validation using Fourier transform spectrometers
Nicholas R. Nalli, Peter J. Minnett, Eric Maddy, W. Wallace McMillan, and Mitchell D. Goldberg
Nicholas R. Nalli, Peter J. Minnett, Eric Maddy, W. Wallace McMillan, and Mitchell D. Goldberg, "Emissivity and reflection model for calculating unpolarized isotropic water surface-leaving radiance in the infrared. 2: Validation using Fourier transform spectrometers," Appl. Opt. 47, 4649-4671 (2008)
The surface-leaving radiance model developed in Part I [Appl. Opt. 47, 3701 (2008)] is validated against an exhaustive set of Fourier transform spectrometer field observations acquired at sea. Unlike prior limited studies, these data include varying all-sky atmospheric conditions (clear, cloudy, and dusty), with regional samples from the tropics, mid-latitudes, and high latitudes. Our analyses show the model to have reduced bias over standard models at emission angles .
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
N are the total number of days having data and passing QA, n are the corresponding number of FTS spectra, are the mean of daily mean wind speeds at (), and are the mean of daily mean retrieved skin SSTs ().
denotes the spectral mean for the LWIR spectral microwindows defined by 980–996, (chosen to avoid lines). Instrument pointing angles were at , except for CSP (as indicated). PD66, HQ73, DW75, S81, W89, BL96, and N05 refer to the published refractive indices [27, 28, 29, 30, 31, 32, 38] used in the model calculations.
Similar to Table 2, except for the highly transparent SWIR narrowband window defined by . A narrow band was chosen, as opposed to a microwindow, to minimize FTS SWIR spectral random noise. Instrument pointing angles were at , except for CSP and ABOVE (as indicated). Note that N05 [36] refractive indices are not available for SWIR channels.
Table 4
Statistical Summary Grouped by and Spectral Regiona
Paired t Test
Wind Speed
Skin SST
95% CI
Window
σ
σ
t
DF
p
Lower
Upper
LWIR
6.2
2.5
28.9
1.2
reject
43
134
0.00
0.05
0.06
LWIR
7.0
3.0
19.8
8.4
reject
83
794
0.00
0.15
0.15
LWIR
6.2
2.5
28.9
1.2
reject
62
134
0.00
0.21
0.22
SWIR
5.4
2.4
26.4
2.9
reject
68
229
0.00
0.04
0.04
SWIR
7.0
3.3
19.7
8.5
reject
100
739
0.00
0.09
0.09
SWIR
6.1
2.7
28.8
1.2
reject
92
119
0.00
0.16
0.17
Paired Student’s t test for the difference in daily mean resulting from the two SLR models for all campaign data and five published refractive indices [27, 28, 29, 31, 32] (omitting S81 [30] and N05 [38]) grouped by emission angle and spectral window; at the 5% significance level, t denotes the test statistic, DF is the degrees of freedom (), p is the probability of surpassing t under , and CI is the 95% confidence interval (defined by the upper and lower bounds). LWIR and SWIR denote the spectral windows defined in Tables 2, 3, respectively. For reference, the means and standard deviations of the wind speed and skin SST are also given.
Similar to Table 4 except showing 95% confidence intervals (CI) and standard deviation for (bias) obtained from each SLR model considered individually.
N are the total number of days having data and passing QA, n are the corresponding number of FTS spectra, are the mean of daily mean wind speeds at (), and are the mean of daily mean retrieved skin SSTs ().
denotes the spectral mean for the LWIR spectral microwindows defined by 980–996, (chosen to avoid lines). Instrument pointing angles were at , except for CSP (as indicated). PD66, HQ73, DW75, S81, W89, BL96, and N05 refer to the published refractive indices [27, 28, 29, 30, 31, 32, 38] used in the model calculations.
Similar to Table 2, except for the highly transparent SWIR narrowband window defined by . A narrow band was chosen, as opposed to a microwindow, to minimize FTS SWIR spectral random noise. Instrument pointing angles were at , except for CSP and ABOVE (as indicated). Note that N05 [36] refractive indices are not available for SWIR channels.
Table 4
Statistical Summary Grouped by and Spectral Regiona
Paired t Test
Wind Speed
Skin SST
95% CI
Window
σ
σ
t
DF
p
Lower
Upper
LWIR
6.2
2.5
28.9
1.2
reject
43
134
0.00
0.05
0.06
LWIR
7.0
3.0
19.8
8.4
reject
83
794
0.00
0.15
0.15
LWIR
6.2
2.5
28.9
1.2
reject
62
134
0.00
0.21
0.22
SWIR
5.4
2.4
26.4
2.9
reject
68
229
0.00
0.04
0.04
SWIR
7.0
3.3
19.7
8.5
reject
100
739
0.00
0.09
0.09
SWIR
6.1
2.7
28.8
1.2
reject
92
119
0.00
0.16
0.17
Paired Student’s t test for the difference in daily mean resulting from the two SLR models for all campaign data and five published refractive indices [27, 28, 29, 31, 32] (omitting S81 [30] and N05 [38]) grouped by emission angle and spectral window; at the 5% significance level, t denotes the test statistic, DF is the degrees of freedom (), p is the probability of surpassing t under , and CI is the 95% confidence interval (defined by the upper and lower bounds). LWIR and SWIR denote the spectral windows defined in Tables 2, 3, respectively. For reference, the means and standard deviations of the wind speed and skin SST are also given.
Similar to Table 4 except showing 95% confidence intervals (CI) and standard deviation for (bias) obtained from each SLR model considered individually.