OSA's Digital Library

Optics Express

Optics Express

  • Editor: Andrew M. Weiner
  • Vol. 21, Iss. 13 — Jul. 1, 2013
  • pp: 15654–15663
« Show journal navigation

Atmospheric corrections of passive microwave data for estimating land surface temperature

Zeng-Lin Liu, Hua Wu, Bo-Hui Tang, Shi Qiu, and Zhao-Liang Li  »View Author Affiliations


Optics Express, Vol. 21, Issue 13, pp. 15654-15663 (2013)
http://dx.doi.org/10.1364/OE.21.015654


View Full Text Article

Acrobat PDF (976 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

Quantitative analysis of the atmospheric effects on observations made by the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) has been performed. The differences between observed brightness temperatures at the top of the atmosphere and at the bottom of the atmosphere were analyzed using a database of simulated observations, which were configured to replicate AMSR-E data. The differences between observed brightness temperatures at the top of the atmosphere and land surface-emitted brightness temperatures were also computed. Quantitative results show that the atmosphere has different effects on brightness temperatures in different AMSR-E channels. Atmospheric effects can be neglected at 6.925 and 10.65 GHz, when the standard deviation is less than 1 K. However, at other frequencies and polarizations, atmospheric effects on observations should not be neglected. An atmospheric correction algorithm was developed at 18.7 GHz vertical polarization, based on the classic split-window algorithm used in thermal remote sensing. Land surface emission can be estimated with RMSE = 0.99 K using the proposed method. Using the known land surface emissivity, Land Surface Temperature (LST) can be retrieved. The RMSE of retrieved LST is 1.17 K using the simulated data.

© 2013 OSA

1. Introduction

Land surface remote sensing satellite observations are affected by both the land surface and the atmosphere. To retrieve land surface parameters from a satellite measurement, the influence of atmosphere must first be removed.

The influence of the atmosphere on the measurement and retrieval of Land Surface Temperature (LST) and Land Surface Emissivity (LSE), using infrared remote sensing data, is obvious and needs to be considered [1

1. F. Nerry, M. P. Stoll, and A. Malaplate, “Multi temporal regression method for mid infrared [3-5µm] emissivity outdoor measurements,” Opt. Express 12(26), 6574–6588 (2004). [CrossRef] [PubMed]

4

4. Z.-L. Li, H. Wu, N. Wang, S. Qiu, J. A. Sobrino, Z. Wan, B.-H. Tang, and G. J. Yan, “Land surface emissivity retrieval from satellite data,” Int. J. Remote Sens. 34(9-10), 3084–3127 (2013). [CrossRef]

]. Compared to optical and infrared remote sensing, passive microwave remote sensing is considered to be an effective tool for all-weather monitoring of land surface processes. The long wavelengths used in passive microwave remote sensing resist atmospheric disturbances and are even able to penetrate clouds and some rainfall. Therefore, after correcting for the atmosphere, passive microwave remote sensing data can be used to retrieve land surface parameters under almost any weather conditions.

In this study, a simulated database was built that covers various atmospheric and land surface conditions and that is configured to reflect AMSR-E data (section 2). Analysis of atmospheric effects on AMSR-E observations and the quantitative results are shown in section 3. An atmospheric correction algorithm at 18.7 GHz vertical polarization is proposed in section 4. Finally, the conclusions are summarized in the last section.

2. Database

A simulated database of passive microwave land surface brightness temperatures, configured to represent AMSR-E data, was built using the modified microwave radiative transfer model (MWMOD) [8

8. R. Fuhrhop and C. Simmer, MWMOD User Manual, Version 1.12 (Institut für Meereskunde, Kiel, Germany, 1998).

], which is used for analyzing the atmospheric effects. To generate the simulated database, atmosphere profile data, LSE data and LST data are all required. Descriptions of the data used in this study are provided below.

2.1 Atmosphere

Atmospheric effects on land surface observations are mainly due to absorption and scatter in the atmosphere. In the microwave portion of the electromagnetic spectrum, water vapor and oxygen are primarily responsible for disrupting observations. Because the wavelength is considerably longer than the size of either water vapor or oxygen molecules, the influence of scatter is usually considered negligible and only absorption need to be considered under clear sky conditions.

There is a special module in the MWMOD to calculate atmospheric parameters associated with a known atmospheric profile. In this atmospheric module, the millimeter-wave propagation model (MPM) is used to describe the propagation characteristics of the path and calculate the atmospheric parameters. Forty-four O2 and 30 H2O local line spectra, which included an empirical water vapor term that can be used to reconcile discrepancies with observed absorptions, were used in this model. To simulate various weather conditions, six types of standard atmospheric profiles in MODTRAN 4, covering a wide range of conditions including bottom of the atmospheric temperature (257.2–299.7 K) and total column precipitable water vapor (4.18–41.92 kg/m2), were used to establish the database in this study. To make the simulated data more representatives, the humidity profiles were adjusted by scaling from 0.5 to 1.5 with a step of 0.1. Then, sixty-six atmospheric profiles covering wider range of humidity condition were generated.

2.2 Soil emissivity

Many studies about soil emission have been performed and several semi-empirical models [9

9. B. J. Choudhury, T. J. Schmugge, A. Chang, and R. W. Newton, “Effect of surface roughness on the microwave emission from soils,” J. Geophys. Res. 84(C9), 5699–5706 (1979). [CrossRef]

,10

10. J. R. Wang and B. J. Choudhury, “Remote sensing of soil moisture content over bare fields at 1.4 GHz frequency,” J. Geophys. Res. 86(C6), 5277–5282 (1981). [CrossRef]

] and physics based models have been developed to describe the emissivity of natural surfaces [11

11. A. K. Fung, Microwave Scattering and Emission Models and Their Applications (Artech House, Boston, 1994).

,12

12. K. S. Chen, T. D. Wu, L. Tsang, Q. Li, J. C. Shi, and A. K. Fung, “Emission of rough surfaces calculated by the integral equation method with comparison to three-dimensional moment method simulations,” IEEE Trans. Geosci. Remote 41(1), 90–101 (2003). [CrossRef]

]. In this study, the Advanced Integral Equation Model (AIEM) [12

12. K. S. Chen, T. D. Wu, L. Tsang, Q. Li, J. C. Shi, and A. K. Fung, “Emission of rough surfaces calculated by the integral equation method with comparison to three-dimensional moment method simulations,” IEEE Trans. Geosci. Remote 41(1), 90–101 (2003). [CrossRef]

] was combined with the Dobson model to simulate the emissivity of a wide range of soil moisture and surface roughness conditions. The surface conditions were configured to reflect data from the AMSR-E instrument, i.e., 6.925, 10.65, 18.7, 23.8, 36.5 and 89 GHz at vertical and horizontal polarization with a 55° incident angle. In the database, the soil surface dielectric constants are described by volumetric soil moisture (sm) that varies from 2% to 44% at a 2% intervals, based on Dobson’s dielectric mixing model for a given soil texture. The surface roughness parameters have a root-mean-square height (s) that ranges from 0.25 to 3.0 cm, with a 0.25 cm interval, and these parameters have a correlation length (l) that ranges from 5 to 30 cm, with a 2.5 cm interval. Figure 1
Fig. 1 Histogram of simulated emissivity.
shows the histogram of simulated emissivity.

2.3 Land surface temperature (LST)

LST is a necessary input parameter to simulate the brightness temperature of a land surface. To make the value of LST more reasonable, LST was set within a given range based on the temperature (T0) of the bottom layer of atmosphere. Specifically, LST varied from −5 K to 15 K, with an interval of 5 K, for T0 ≥ 280 K and from −10 K to 10 K, with an interval of 5 K, for T0<280 K.

Using this input data, a simulation database containing 638880 AMSR-E brightness temperature observations and covering a range of atmospheric and land surface conditions was established using the MWMOD. All of following analysis work is based on the simulated data.

3. Analysis of atmospheric effects

To show the effects of atmosphere on AMSE-E observations, the simulated brightness temperature at the top and the bottom of the atmosphere (TBp_toa and TBp_boa) has been compared to the land surface emitted brightness temperature (TBp_land). According to radiative transfer theory, the satellite-observed brightness temperature at polarization mode, p, and a given frequency and incidence angle can be written as:
TBp_toa=TBp_boa×t+Tau,
(1)
where Tau is the upwelling atmospheric emission, t is atmospheric transmittance and TBp_boa is the brightness temperature at polarization, p, (V-vertical or H-horizontal) observed at the bottom of atmosphere. It can be expressed as:
TBp_boa=TBp_land+[Tad+Tsky×t]×(1ep),
(2)
and
TBp_land=Ts×ep,
(3)
where Ts is the physical temperature of the land surface and ep is the land surface emissivity at polarization, p. Tad is the downwelling atmospheric emission and Tsky is the space background brightness temperature, which is approximately 2.7 K and is usually neglected.

Figure 2
Fig. 2 Comparison between the TBp_toa and TBp_boa at vertical polarization (a) and horizontal polarization (b).
shows the difference between TBp_toa and TBp_boa at vertical (△TBv) and horizontal (△TBH) polarization using simulated data for all of the AMER-E channels. The mean and standard deviation (STD) for the difference between △TBv and △TBH are given in Table 1

Table 1. The mean and standard deviation (STD) of the △TBv and △TBH for all channels of AMSR-E.

table-icon
View This Table
| View All Tables
.

The simulated data set shows that the mean differences are greater than zero, with the exception of △TB89v. This means that the atmosphere acts as an emission source rather than an absorption layer in most cases. The STD resulting from a comparison of TBp_toa and TBp_boa is 0.44 K and 0.65 K, respectively, at a frequency of 6.925 GHz for the vertical and horizontal polarizations. This means that the atmospheric effect on observed brightness temperatures at 6.925 GHz can be neglected and the TB6.925p_toa can be directly used as TB6.925p_boa without introducing significant errors. Furthermore, because the mean value is 0.6 K and the STD is 0.57 K, TB10.65v_toa can be used as TB10.65v_boa without considering atmospheric effects. For the other channels of AMSR-E, the mean and STD are larger than 1 K, with the exception of the STD at 10.65 GHz with horizontal polarization and the mean at 36.5 GHz with vertical polarization. Atmospheric effects on these channels should be considered.

Figure 3
Fig. 3 Comparison between the TBp_toa and TBp_land at vertical polarization (a) and horizontal polarization (b).
shows the difference between TBp_toa and TBp_land at vertical (△T’Bv) and horizontal (△T’BH) polarization using simulated data for all of the channels of AMER-E. The mean and STD of the difference between △T’Bv and △T’BH are given in Table 2

Table 2. The mean and STD of the △T’Bv and △T’BH for all channels of AMSR-E.

table-icon
View This Table
| View All Tables
.

Comparing TBp_toa and TBp_land shows that both the mean and the STD are larger than 1 K in every AMSR-E channel. This means that atmospheric effects should be considered when estimating TBp_land using AMSR-E data.

Further analysis shows that atmospheric effects are less critical to the temperature of the atmosphere than to the atmospheric water content across a range of atmospheric conditions. A comparison of TBp_toa and TBp_boa shows that, when the atmospheric water content is less than 20 kg/m2, the mean and STD are less than 1.5 K at 18.7 GHz vertical polarization. At 36.5 GHz vertical polarization, the mean and STD are less than 2 K. At 89 GHz vertical polarization, the mean and STD are less than 2 K when the atmospheric water content is less than 10 kg/m2. Due to the low values of land surface emissivity at horizontal polarization, atmospheric effects are more pronounced than at vertical polarization. The results show that atmospheric effects on observed brightness temperatures at horizontal polarization should be considered.

4. Atmospheric correction algorithm

4.1 Algorithm

The results above show that atmospheric correction is necessary for AMSR-E data, with the exception of the 6.925 and 10.65 GHz bands at vertical polarization and 6.925 GHz at horizontal polarization. Using the simulated data, a theoretical description of the required atmospheric correction and a correction algorithm are presented in this section. The approximate equivalency of ev18.7 and ev23.8 [13

13. F. Z. Weng and N. C. Grody, “Physical retrieval of land surface temperature using the special sensor microwave imager,” J. Geophys. Res. 103(D8), 8839–8848 (1998). [CrossRef]

], as well as the obvious differences between the atmospheric effects at 18.7 and 23.8 GHz vertical polarization, makes it possible to remove the atmospheric effects on observations at 18.7 GHz vertical polarization. Based on the classic split window technique, which is usually used in thermal infrared remote sensing [14

14. F. Becker and Z.-L. Li, “Towards a local split window method over land surfaces,” Int. J. Remote Sens. 11(3), 369–393 (1990). [CrossRef]

], an atmospheric correction method for the 18.7 GHz vertical polarization channel of AMSR-E was established.

Tad=Tau+2.11,R2=0.99.
(4)

Equation (4) demonstrates that it is reasonable to consider Tau and Tad to be approximately equal [15

15. E. G. Njoku and L. Li, “Retrieval of land surface parameters using passive microwave measurements at 6-18 GHz,” IEEE Trans. Geosci. Remote 37(1), 79–93 (1999). [CrossRef]

], especially under relatively moist atmospheric conditions.

Next, some linear relationships between t and w (the atmospheric water content, kg/m2) at 18.7 and 23.8 GHz were also established:

t18.7=0.003×w+0.975,R2=0.99,
(5)
t23.8=0.007×w+0.951,R2=0.99.
(6)

Figure 4
Fig. 4 Linear relationships between w and t at 18.7 and 23.8 GHz.
shows a scatter plot of t18.7 and t23.8 with w.

Using the approximate expression for effective radiating temperature (Tae), Tau can be expressed as [15

15. E. G. Njoku and L. Li, “Retrieval of land surface parameters using passive microwave measurements at 6-18 GHz,” IEEE Trans. Geosci. Remote 37(1), 79–93 (1999). [CrossRef]

]:

Tau=Tae×(1t).
(7)

Substituting Eqs. (4)-(7) into Eqs. (1) and (2) gives:
TB18.7v_toa=Ts+(TaeTs)×(1b18.7)a18.7×w×(TaeTs)2a18.7×b18.7×w×Ts+r18.7V×a18.7×w×(TaeTs)×(12b18.7)a18.72×w2×r18.7V×Tae+r18.7V×b18.7×(1b18.7)×(TaeTs)sb18.72×r18.7V×T+2.11×r18.7V×(a18.7×w+b18.7),
(8)
TB23.8v_toa=Ts+(TaeTs)×(1b23.8)a23.8×w×(TaeTs)2a23.8×b23.8×w×Ts+r23.8V×a23.8×w×(TaeTs)×(12b23.8)a23.82×w2×r23.8V×Tae+r23.8V×b23.8×(1b23.8)×(TaeTs)b23.82×r23.8V×Ts+2.11×r23.8V×(a23.8×w+b23.8),
(9)
where rp is the soil reflectivity (related to the soil emissivity, ep, by ep = 1- rp).

Assuming that ev18.7ev23.8 and ignoring some small terms in Eqs. (8) and (9), TB18.7v_land can be approximately expressed, by combining Eqs. (8) and (9), as:
TB18.7v_land=A1×TB18.7v_toa+A2×(TB18.7v_toaTB23.8v_toa),
(10)
where Ai is an unknown coefficient.

To improve the accuracy of Eq. (10), correction terms were added, and the coefficients were determined from the simulated data.

TB18.7v_land=TB18.7v_toa+0.506×(TB18.7v_toaTB23.8v_toa)0.019×(TB18.7v_toaTB23.8v_toa)20.085.
(11)

Figure 5
Fig. 5 The difference between the simulated values of TB18.7v_toa and TB18.7v_land. △T1 is the difference between the simulated values of TB18.7v_toa and TB18.7v_land.T2 is the difference between the estimated values of TB18.7v_land using Eq. (11) and TB18.7v_land.
shows the difference between TB18.7v_toa and TB18.7v_land (△T1) and the difference between TB18.7v_land estimated by Eq. (11) and TB18.7v_land in the simulated database (△T2).

Figure 5 shows that that this algorithm can correct for the atmospheric effects and that RMSE may decrease from 6.04 K to 0.99 K. Using this algorithm, the land surface emission can be estimated without using any other auxiliary data. Moreover, if e18.7v is known, LST can be estimated from AMSR-E data using the following equation.

Ts=TB18.7v_toae18.7v+0.506×(TB18.7v_toaTB23.8v_toa)e18.7v0.019×(TB18.7v_toaTB23.8v_toa)2e18.7v0.085e18.7v.
(12)

Using Eq. (12), the estimated value of LST from the simulated data is shown in Fig. 6
Fig. 6 The difference between Ts estimated by Eq. (12) and the actual value of Ts.
. The RMSE of the estimated LST is 1.17 K.

4.2 Sensitivity analysis

Partial derivatives, with respect to TB18.7v_toa, TB23.8v_toa and e18.7v, were computed from Eq. (12). Respectively, they are:

TsTB18.7v_toa=1.5060.038×(TB18.7v_toaTB23.8v_toa)e18.7v,
(13)
TsTB23.8v_toa=0.506+0.038×(TB18.7v_toaTB23.8v_toa)e18.7v,
(14)
Tse18.7v=TB18.7v_toa+0.506×(TB18.7v_toaTB23.8v_toa)0.019×(TB18.7v_toaTB23.8v_toa)2e18.7v2.
(15)

Figure 7
Fig. 7 Histograms of the results of Eqs. (13) and (14) using the simulated data.
shows the results of Eqs. (13) and (14), computed using the simulated data. The results show that, if △TB18.7v_toa (△TB23.8v_toa) is 1 K, △Ts will be approximately 1.98 K (−0.85 K). Figure 8
Fig. 8 Histogram of the results computed by Eq. (15), based on the simulated data.
shows the data based on the calculation in Eq. (15). The bias and STD of Eq. (15) are −325.41 K and 32.77 K, respectively. This means that, if △e18.7v is 0.01, △Ts will be approximately −3.25 K.

5. Conclusions

Using the simulated database, which covers a wide range of atmospheric and land surface conditions, a quantitative analysis of the atmospheric effects on the AMSR-E data was shown in this study. The results show that the atmosphere has an obvious effect on observations of AMSR-E, with the exception of the 6.925 and 10.65 GHz bands at vertical polarization and 6.925 GHz at horizontal polarization. Atmospheric correction is necessary before AMSR-E data can be used to estimate land surface parameters. Using empirical relationships and reasonable assumptions, an atmospheric correction algorithm is proposed. Based on observations of two AMSR-E channels (18.7 and 23.8 GHz vertical polarization), atmospheric effects on observations at 18.7 GHz vertical polarization can be removed effectively and the emission of land surfaces, at 18.7 GHz vertical polarization, can be estimated. Additionally, LST can also be estimated using the proposed method and a known emissivity. The results show that the correction algorithm performs well. With simulated data, the RMSE decreases from 6.04 K to 0.99 K, and the RMSE of the estimated LST was 1.17 K. These findings indicate that the proposed method is helpful for improving estimates of land surface parameters.

Acknowledgments

This work was supported by the National Natural Science Foundation of China under Grants 41101332 and 41231170, the “Strategic Priority Research Program” of the Institute of Geographic Science and Nature Resources Research (IGSNRR) and the Chinese Academy of Sciences (CAS) under Grant 2012SJ004.

References and links

1.

F. Nerry, M. P. Stoll, and A. Malaplate, “Multi temporal regression method for mid infrared [3-5µm] emissivity outdoor measurements,” Opt. Express 12(26), 6574–6588 (2004). [CrossRef] [PubMed]

2.

X. Y. OuYang, N. Wang, H. Wu, and Z.-L. Li, “Errors analysis on temperature and emissivity determination from hyperspectral thermal infrared data,” Opt. Express 18(2), 544–550 (2010). [CrossRef] [PubMed]

3.

Z.-L. Li, B.-H. Tang, H. Wu, H. Ren, G. J. Yan, Z. Wan, I. F. Triggo, and J. A. Sobrino, “Satellite-derived land surface temperature: Current status and perspectives,” Remote Sens. Environ. 131, 14–37 (2013). [CrossRef]

4.

Z.-L. Li, H. Wu, N. Wang, S. Qiu, J. A. Sobrino, Z. Wan, B.-H. Tang, and G. J. Yan, “Land surface emissivity retrieval from satellite data,” Int. J. Remote Sens. 34(9-10), 3084–3127 (2013). [CrossRef]

5.

C. Prigent, W. B. Rossow, and E. Matthews, “Microwave land surface emissivities estimated from SSM/I observations,” J. Geophys. Res. 102(D18), 21867–21890 (1997). [CrossRef]

6.

F. Karbou, C. Prigent, L. Eymard, and J. R. Pardo, “Microwave land emissivity calculations using AMSU measurements,” IEEE Trans. Geosci. Remote 43(5), 948–959 (2005). [CrossRef]

7.

J. B. Snider, E. R. Westwater, and L. S. Fedor, “Radiometric correction for atmospheric effects in surface sensing from aircraft and satellites,” in Passive Microwave Remoter Sensing of Land-Atmosphere Interactions, E.D. B.J. Choudhury, Y. H. Kerr, and P. Pampaloni, ed. (VSP, Zeist, Netherlands, 1994).

8.

R. Fuhrhop and C. Simmer, MWMOD User Manual, Version 1.12 (Institut für Meereskunde, Kiel, Germany, 1998).

9.

B. J. Choudhury, T. J. Schmugge, A. Chang, and R. W. Newton, “Effect of surface roughness on the microwave emission from soils,” J. Geophys. Res. 84(C9), 5699–5706 (1979). [CrossRef]

10.

J. R. Wang and B. J. Choudhury, “Remote sensing of soil moisture content over bare fields at 1.4 GHz frequency,” J. Geophys. Res. 86(C6), 5277–5282 (1981). [CrossRef]

11.

A. K. Fung, Microwave Scattering and Emission Models and Their Applications (Artech House, Boston, 1994).

12.

K. S. Chen, T. D. Wu, L. Tsang, Q. Li, J. C. Shi, and A. K. Fung, “Emission of rough surfaces calculated by the integral equation method with comparison to three-dimensional moment method simulations,” IEEE Trans. Geosci. Remote 41(1), 90–101 (2003). [CrossRef]

13.

F. Z. Weng and N. C. Grody, “Physical retrieval of land surface temperature using the special sensor microwave imager,” J. Geophys. Res. 103(D8), 8839–8848 (1998). [CrossRef]

14.

F. Becker and Z.-L. Li, “Towards a local split window method over land surfaces,” Int. J. Remote Sens. 11(3), 369–393 (1990). [CrossRef]

15.

E. G. Njoku and L. Li, “Retrieval of land surface parameters using passive microwave measurements at 6-18 GHz,” IEEE Trans. Geosci. Remote 37(1), 79–93 (1999). [CrossRef]

OCIS Codes
(280.0280) Remote sensing and sensors : Remote sensing and sensors
(300.6370) Spectroscopy : Spectroscopy, microwave
(010.1285) Atmospheric and oceanic optics : Atmospheric correction
(280.4991) Remote sensing and sensors : Passive remote sensing

ToC Category:
Remote Sensing

History
Original Manuscript: May 22, 2013
Revised Manuscript: June 17, 2013
Manuscript Accepted: June 17, 2013
Published: June 21, 2013

Citation
Zeng-Lin Liu, Hua Wu, Bo-Hui Tang, Shi Qiu, and Zhao-Liang Li, "Atmospheric corrections of passive microwave data for estimating land surface temperature," Opt. Express 21, 15654-15663 (2013)
http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-21-13-15654


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. F.  Nerry, M. P.  Stoll, A.  Malaplate, “Multi temporal regression method for mid infrared [3-5µm] emissivity outdoor measurements,” Opt. Express 12(26), 6574–6588 (2004). [CrossRef] [PubMed]
  2. X. Y.  OuYang, N.  Wang, H.  Wu, Z.-L.  Li, “Errors analysis on temperature and emissivity determination from hyperspectral thermal infrared data,” Opt. Express 18(2), 544–550 (2010). [CrossRef] [PubMed]
  3. Z.-L.  Li, B.-H.  Tang, H.  Wu, H.  Ren, G. J.  Yan, Z.  Wan, I. F.  Triggo, J. A.  Sobrino, “Satellite-derived land surface temperature: Current status and perspectives,” Remote Sens. Environ. 131, 14–37 (2013). [CrossRef]
  4. Z.-L.  Li, H.  Wu, N.  Wang, S.  Qiu, J. A.  Sobrino, Z.  Wan, B.-H.  Tang, G. J.  Yan, “Land surface emissivity retrieval from satellite data,” Int. J. Remote Sens. 34(9-10), 3084–3127 (2013). [CrossRef]
  5. C.  Prigent, W. B.  Rossow, E.  Matthews, “Microwave land surface emissivities estimated from SSM/I observations,” J. Geophys. Res. 102(D18), 21867–21890 (1997). [CrossRef]
  6. F.  Karbou, C.  Prigent, L.  Eymard, J. R.  Pardo, “Microwave land emissivity calculations using AMSU measurements,” IEEE Trans. Geosci. Remote 43(5), 948–959 (2005). [CrossRef]
  7. J. B. Snider, E. R. Westwater, and L. S. Fedor, “Radiometric correction for atmospheric effects in surface sensing from aircraft and satellites,” in Passive Microwave Remoter Sensing of Land-Atmosphere Interactions, E.D. B.J. Choudhury, Y. H. Kerr, and P. Pampaloni, ed. (VSP, Zeist, Netherlands, 1994).
  8. R. Fuhrhop and C. Simmer, MWMOD User Manual, Version 1.12 (Institut für Meereskunde, Kiel, Germany, 1998).
  9. B. J.  Choudhury, T. J.  Schmugge, A.  Chang, R. W.  Newton, “Effect of surface roughness on the microwave emission from soils,” J. Geophys. Res. 84(C9), 5699–5706 (1979). [CrossRef]
  10. J. R.  Wang, B. J.  Choudhury, “Remote sensing of soil moisture content over bare fields at 1.4 GHz frequency,” J. Geophys. Res. 86(C6), 5277–5282 (1981). [CrossRef]
  11. A. K. Fung, Microwave Scattering and Emission Models and Their Applications (Artech House, Boston, 1994).
  12. K. S.  Chen, T. D.  Wu, L.  Tsang, Q.  Li, J. C.  Shi, A. K.  Fung, “Emission of rough surfaces calculated by the integral equation method with comparison to three-dimensional moment method simulations,” IEEE Trans. Geosci. Remote 41(1), 90–101 (2003). [CrossRef]
  13. F. Z.  Weng, N. C.  Grody, “Physical retrieval of land surface temperature using the special sensor microwave imager,” J. Geophys. Res. 103(D8), 8839–8848 (1998). [CrossRef]
  14. F.  Becker, Z.-L.  Li, “Towards a local split window method over land surfaces,” Int. J. Remote Sens. 11(3), 369–393 (1990). [CrossRef]
  15. E. G.  Njoku, L.  Li, “Retrieval of land surface parameters using passive microwave measurements at 6-18 GHz,” IEEE Trans. Geosci. Remote 37(1), 79–93 (1999). [CrossRef]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited