SEARCH

SEARCH BY CITATION

Keywords:

  • microwave radiometers;
  • soil moisture;
  • land surface emissivity;
  • aircraft measurements;
  • SSM/I

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[1] Aircraft microwave radiometer measurements at 19 and 37 GHz were made over a 40 km2 agricultural area in southern Ontario on 16 and 23 May 2001 when the average soil moisture in the area was 28 and 40%, respectively. SSM/I satellite data and ground-based measurements of soil moisture were collected over the period 2 May to 25 July 2001. The emissivity of a water body (Lake Huron) calculated from the aircraft and satellite microwave radiometer measurements agreed with model calculations to within 0.02, except for the aircraft 37 GHz V channel. The 19 GHz emissivity measured by the SSM/I was higher than coincident aircraft measurements by 0.02 in moister soil conditions and 0.04 in drier conditions. Vegetation height increased from a maximum of 40 cm in hay fields in May to up to 200 cm in corn fields in July. There was a statistically significant relationship between soil moisture and 19 GHz H data for aircraft and SSM/I measurements in May. However, the standard error in the soil moisture estimate was 7%. Soil moisture seemed to have very little influence on 19 and 37 GHz emission during June and July. When the mean monthly emissivity for the May to July period was calculated from the SSM/I data, there was found to be no significant variation from month to month.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[2] The measurement of soil moisture using remotely sensed techniques is of increasing interest in Canada due to the impracticality of measuring soil moisture using in situ methods over such a large country and the importance of soil moisture information for hydrological purposes, meteorological forecasts and climate modeling. Because the soil depth contributing to a passive microwave signal is 0.06 to 0.2 wavelengths [Wilheit, 1978; Wang, 1987], future data from sensors such as the AMSR (Advanced Microwave Scanning Radiometer) with a 6.9 GHz channel are much more suited to this purpose than the current SSM/I (Special Sensor Microwave Imager) which has a lowest-frequency channel at 19.35 GHz. On the other hand, no satellite observations at frequencies lower than 19.35 GHz have been made since 1981–1987 when the SMMR (Scanning MultiChannel Radiometer) acquired 6.6 GHz data. The question of whether soil moisture information can be obtained from the SSM/I 19.35 GHz channel is therefore of interest from a climatic point of view and a study by Jackson [1997] indicated that this may be possible. Another study by Vinnikov et al. [1999], used the SMMR 18 GHz channel to demonstrate that it may be a useful source of soil moisture information.

[3] There is also interest in characterizing land surface emissivity at higher frequencies in order to fully exploit the atmospheric sounding capability of the satellite over land. A number of studies of land surface emissivity have already been carried out [e.g., Jones and Vonder Haar, 1997; Prigent et al., 1997; Morland et al., 2001; Bennartz et al., 2002].

[4] This paper presents aircraft and SSM/I 19 and 37 GHz observations obtained over an agricultural area in southern Ontario during the spring growth period. The data sets are analyzed firstly to investigate the ability of the SSM/I lower-frequency channels to provide information on soil moisture in this type of area and secondly to investigate the variability of microwave emissivity.

2. Field Campaign and Aircraft Data

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[5] The Meteorological Service of Canada (MSC) and the National Research Council of Canada (NRC) have jointly developed an aircraft research facility for conducting airborne research to address atmospheric and climate science issues [MacPherson et al., 2001]. The Twin Otter aircraft based at the NRC in Ottawa is equipped with 19, 37 and 85 GHz radiometers that are side-viewing at a fixed angle of 53°. At the flying height of 300 m chosen for this experiment, the 3 dB footprint is 87 m across track by 52 m along track.

[6] On 16 and 23 May 2001, the Twin Otter aircraft made measurements of microwave emission along a number of flight lines in an agricultural area west of Toronto (see Figure 1). The area was chosen because of its flat topography and relatively large field sizes. Typical fields were 600 to 1000 m in length in a north–south direction and 400 to 600 m in width in an east–west direction. For each flight line, the flight path was chosen so that the center of the instrument footprint was about 200 m from the edge of the fields. Side roads intersected the main roads at approximately 2 km intervals. On the flight days capacitance and gravimetric techniques were used to conduct in situ sampling of soil moisture in the fields on both sides of each side road at the approximate center of the microwave footprint. Ten capacitance probe measurements and one gravimetric measurement were made at each sample point. The 6 cm long capacitance probe measures average soil moisture over 0 to 6 cm depth, while the soil samples for gravimetric measurement were obtained over 0 to 10 cm depth. Bulk density measurements were used to convert the gravimetric measurements to volumetric soil moisture. The average difference between the gravimetric and capacitance probe observations was +0.4% soil moisture by volume, with a standard deviation of 5.5% soil moisture by volume.

image

Figure 1. Top left map shows the agricultural area studied (small box to east of Lake Huron) and the area of Lake Huron used to check the satellite measurements. Crosses mark the positions of SSM/I 19 GHz pixel centers during the 1400 overpass on 23 April 2001. Bottom right map shows the positions of the flight lines in the agricultural study area as well as the TDR probe locations.

Download figure to PowerPoint

[7] In order to obtain a time series of soil moisture measurements, particularly for the satellite data analysis, time domain reflectometry (TDR) probes were installed at four different field sites (see Figure 1). Soil moisture measurements were acquired at 5 and 10 cm depths at three hourly intervals from 2 May to 25 July 2001. The theoretical relationship relating time delay to soil moisture, results in larger errors in clay soils [Hook and Livingston, 1995]. The soil type in the study area was mainly silty clay loam. To assess the performance of the TDR soil moisture measurement at the various sites, gravimetric measurements were made close to the TDR probes on seven different occasions. In order to compare the 0 to 10 cm gravimetric measurement, the TDR measurement was taken to be the average of the 5 and 10 cm TDR probe measurements. The TDR measurements were between 2 and 16% soil moisture by volume higher than the gravimetric measurement with the difference depending on the site. The r2 correlation coefficients were 0.77 to 0.93. This comparison was used to correct the average TDR measurement.

[8] Figure 2 shows the average raw and gravimetrically adjusted TDR measurements for all four sites compared to the average gravimetric measurement along the flight lines. Fifty-six gravimetric measurements were made on the 16 May and 63 on the 23 May. The differences between them are 4.0% and 0.2% soil moisture by volume on the 16 and 23 May, respectively. These relatively small differences indicate that the TDR probes are able to give a fairly good estimate of the areal soil moisture.

image

Figure 2. Average 5 and 10 cm time domain reflectometry (TDR) probe measurement before and after adjustment with nearby gravimetric measurements. Also presented is the average of the 2 km gravimetric measurements made along the flight lines on 16 and 23 May.

Download figure to PowerPoint

[9] A weather station at one of the TDR sites (see Figure 1) operated from 2 May to 26 July 2001 and monitored a number of meteorological variables including soil temperature at 5 and 10 cm depths, temperature at 2.5 cm above the soil surface, air temperature, precipitation and humidity.

[10] NDVI values for the area were calculated from the AVHRR (Advanced Very High Resolution Radiometer) polar orbiting instrument channel 1 (0.58 to 0.68 μm) and channel 2 (0.73 to 1.1 μm) data. NDVI values were also calculated from the aircraft Exotech 100BX instrument [MacPherson et al., 2001], based on the difference between observations at 597 to 700 μm and 804 to 1045 μm. NDVI calculated from the AVHRR instrument is compared to that calculated from the aircraft measurements in Figure 3. In mid-May, only about 40% of the fields were vegetated and of those that were, average vegetation height was 15 cm with a maximum height of around 40 cm in hay fields. By the end of July, all fields were vegetated with a maximum height of around 2 m in corn fields. These changes are reflected in the increasing NDVI.

image

Figure 3. Average NDVI and standard deviation (y error bars) for the study area calculated from AVHRR satellite data and from the aircraft instrument.

Download figure to PowerPoint

3. Satellite Data

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[11] Archived 19 and 37 GHz SSM/I brightness temperatures were obtained from the Canadian Meteorological Centre for the period during which the TDR probes operated, 2 May to 26 July 2001. At these frequencies the SSM/I has a spatial resolution of 43 × 69 km and 28 × 37 km, respectively [Hollinger et al., 1990]. The data values for the SSM/I pixel centers falling within the agricultural study area were averaged to obtain a measurement for each overpass. The western edge of the study area was at least 25 km from Lake Huron. Since the SSM/I 19 GHz resolution is 43 km in a cross track direction and 69 km along track, the pixels included in the analysis were assumed to not be affected by the lake boundary. In order to check the calculation of microwave emissivity, SSM/I data were also averaged over a one degree square area of Lake Huron, also shown in Figure 1.

[12] Cloud cover was identified using 5 km resolution GOES-8 channel 1 and channel 4 data obtained from the Canadian Meteorological Centre archive. Three hourly data were available for May and June 2001 and hourly data for July 2001. Because the GOES data were relatively infrequent, the images preceding and following each SSM/I overpass were examined. If there was a possibility of cloud in the study area at the time of the overpass, the SSM/I data were excluded from the emissivity calculation.

4. Calculation of Emissivity

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[13] Microwave emissivity is calculated from either aircraft or satellite observed brightness temperatures as follows.

  • equation image

[14] BTcor is the observed microwave brightness temperature corrected for atmospheric effects between the ground and the sensor. BTatmp[DOWNWARDS ARROW] is the downwelling atmospheric brightness temperature and Tsurface is the physical temperature of the surface.

[15] A microwave radiative transfer model was used to calculate BTcor from the microwave observations as well as BTatmp[DOWNWARDS ARROW]. Absorption by water vapor and oxygen were calculated using the Liebe [1989] and Liebe et al. [1993] models, respectively. Input atmospheric information for the satellite data set was Canadian Meteorological Centre (CMC) model reanalysis temperature and humidity profiles. The CMC data are produced at 00, 06, 12 and 18 UTC. The surface layer in the profiles was modified with ground level pressure, temperature and humidity observations. For the calculation of atmospheric corrections on the aircraft flight days, a radiosonde was launched at the time of the flight from the Centre for Atmospheric Research (CARE) (see Figure 1).

[16] The surface temperature values used in the emissivity calculation for the SSM/I observations were calculated from GOES-8 imager channel 4 observations (10.2 to 11.2 μm). Observations made by the aircraft side-looking infrared radiometer (9.25 to 12.0 μm) provided surface temperature information for the aircraft data set. The method for infrared atmospheric correction is described in Morland et al. [2001]. It was assumed that the agricultural area had an infrared emissivity of 0.965, a value estimated for an agricultural area in Quebec [Goita et al., 1996]. A comparison of 300 points indicated that on average the land surface temperature estimated from the GOES data was 0.9 K lower than the weather station temperature measured just above the soil surface. A linear regression between the two data sets produced an r2 correlation coefficient of 0.84.

[17] In order to check the microwave observations and emissivity calculation, water brightness temperature measurements were made at various view angles during flights over Lake Huron on 16 and 23 May 2001. The emissivity calculated from the aircraft observations was compared with that calculated from a geometric optics model [Petty and Katsaros, 1994]. The inputs to the model were wind speed and surface temperature measured by the aircraft at 30 m above the water surface. Table 1 shows that the average aircraft and model emissivity agree to within 0.02 with the exception of the 37 GHz V channel, which is likely the result of a slight polarization misalignment in the aircraft 37 GHz microwave radiometer.

Table 1. Emissivity Calculated From a Geometric Optics Model Compared With That Calculated From Aircraft and SSM/I Observations Over Lake Huron
ChannelAverage Model Minus Aircraft EmissivityStandard DeviationAverage Model Minus Satellite EmissivityStandard Deviation
19 H−0.0050.021−0.0170.02
19 V0.0080.015−0.0020.012
22 V  −0.0320.024
37 H−0.0130.029−0.0030.022
37 V0.0450.028+0.0160.013
85 H  0.0090.052
85 V  0.0330.023

[18] The emissivity of the area in Lake Huron shown on Figure 1, was calculated from SSM/I observations. Water temperature and wind speed measured by two buoys marked on Figure 1 were used as input to the geometric optics model. The emissivity calculated from the observations and that calculated from the model for cloud-free days are compared in Table 1 and agree on average to within 0.02 for 19 and 37 GHz data. There is a bigger discrepancy between model calculations and 22 and 85 GHz observations. Since these channels are more sensitive to the atmosphere, these results indicate that the atmospheric state was not always well described by the CMC profiles.

5. Aircraft Analysis

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[19] During the aircraft campaign, three larger fields were selected for detailed measurements: a bare field which had been ploughed and raked smooth, an untilled field which was planted with corn around the beginning of May, and a hay field. Vegetation water content at the pasture site increased from 0.72 g kg−1 to 1.68 g kg−1between 16 and 23 May. Vegetation water content at the corn field increased from 0.05 to 0.19 g kg−1 between 16 and 23 May and vegetation height increased from 6 to 12 cm. There was 1 or 2% stubble cover on the bare field and about 30% on the corn field. Standard deviation of surface height was 1.3 cm at both the corn and bare fields and varied between 1.0 and 4.3 cm for fields in the area. Soils in the area were estimated to be silty clay loam from a visual analysis of the texture. Average bulk density was 1.32, 1.35 and 1.4 g m−3 at the bare soil, corn and pasture fields, respectively.

[20] In addition to the 2 to 4 TDR probes installed at each of the field sites, soil moisture was sampled at 200 m intervals across each field site on the flight days using either capacitance or gravimetric techniques. This resulted in approximately sixteen soil moisture samples per field. Soil moisture was higher on 23 than on 16 May due to a rainfall event on 22 May during which four rain gauges installed in the experimental area recorded an average of 25 mm rain.

[21] Figure 4 shows that 19 GHz H emissivity measured by the aircraft radiometers decreased at all three sites between 16 and 23 May with the increase in soil moisture. However, the 37 GHz H emissivity showed much smaller changes. The decrease in the 19 GHz H emissivity is 0.04, 0.06 and 0.07 for the pasture, corn and bare fields, respectively. The emissivity on 16 May (drier soil conditions) is 0.94, 0.89 and 0.87 for the pasture, corn and bare fields. The highest emissivity occurs for the most highly vegetated field, which is in line with previous measurements [Morland et al., 2000].

image

Figure 4. Emissivity observed at three field sites on 16 and 23 May 2001. Soil moisture was sampled at 200 m intervals throughout each field.

Download figure to PowerPoint

[22] The soil moisture measurements made at 2 km intervals along two of the flight tracks were matched to the closest aircraft measurements. Fields were assumed to be bare, mixed or vegetated on the basis of NDVI values calculated from aircraft observations of less than 0.4, between 0.4 and 0.6 and greater than 0.6, respectively. There is some uncertainty in the field classification due to the fact that the instrument used to calculate NDVI was nadir looking, whereas the microwave radiometers had a side-viewing angle of 53°.

[23] The results of a linear regression between soil moisture and brightness temperature and emissivity are shown in Table 2. There is a statistically significant relationship between soil moisture and 19 GHz H brightness temperature, but the standard error in the soil moisture estimate is 7%. This is perhaps not surprising given that Duke et al. [1999] found a similar standard error in predicting soil moisture from 1.4 GHz brightness temperature when variations in soil type and surface roughness were not accounted for. Converting to emissivity results in a weaker relationship with soil moisture.

Table 2. Soil Moisture Regressed Against 19 GHz Brightness Temperature (BT) and Emissivity (E) From Aircraft Observations
Data TypeNDVISlopeInterceptStandard Error in SlopeStandard Error in Soil Moisture,% by volr2Number of PointsSignificant at 5% Level?
BT<0.4−0.61940.170.5431yes
BT0.4 to 0.6−0.72120.290.4815no
BT>0.6−0.41510.170.4719yes
E<0.4−1721805090.2731no
E0.4 to 0.6−12614376110.1615no
E>0.6−1231403970.3419no

6. Comparison of Aircraft and Satellite Data

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[24] In Figure 5, the average emissivity over each of the aircraft flight lines was compared to the average emissivity calculated from satellite data. For both satellite and aircraft data sets the atmospheric correction was carried out using radiosonde observations. On 16 May, observations at London Airport indicated altocumulus at 3.3 km. The microwave radiative transfer model (RTM) was run assuming 0.01 mm cloud between 3 and 4 km. On 23 May, there was patchy cumulus present and the RTM was run assuming 0.01 mm cloud between 1 and 2 km.

image

Figure 5. Average satellite (Sat) emissivity for the study area compared to aircraft (Air) emissivity averaged along each flight line for 16 and 23 May 2001.

Download figure to PowerPoint

[25] Figure 5 shows that the 19 GHz H satellite emissivity is higher than the aircraft emissivity by 0.04 and 0.02 in dry and wetter conditions, respectively. The 37 GHz H satellite emissivity is 0.03 higher than the aircraft emissivity during dry conditions and similar in moist soil conditions. Since the same atmospheric information was used in the correction, this might be due to the area observed by the satellite being more highly vegetated than that represented by aircraft flight tracks. The aircraft radiometers sampled only about 2 or 3% of the SSM/I footprint and flight lines were close to roads. Open farmland tended to be closer to roads and woodlots tended to lie behind the fields. On the other hand, Figure 3 shows that although the aircraft NDVI is more variable, it tends to be higher than the AVHRR value.

7. Satellite Analysis

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[26] As a preliminary check, soil moisture was regressed against SSM/I normalized brightness temperature for the months of May, June and July separately. No cloud masking was carried out on this initial data set. Table 3 shows how the slope of the relationship and the r2 correlation coefficient decrease as the season progresses and vegetation cover increases. A T-test indicates that the relationship is significant at the 5% level in May but not during other months.

Table 3. Soil Moisture and 19 GHz H Brightness Temperature (BT) and Emissivity (E) From SSM/I Observationsa
MonthDate TypeSlopeInterceptStandard Error in SlopeStandard Error in Soil Moisture,% by volr2Number of PointsSignificant at 5% Level?
  • a

    Emissivity data set is cloud masked, but brightness temperature data set is not.

MayBT−0.31030.0560.23110Yes
JuneBT−0.2770.0340.17128No
JulyBT0.0290.0540.01103No
MayE−801042350.3225No
JuneE−67952140.2432No
JulyE−29421730.0928No

[27] It might be thought that there would be a stronger relationship between soil moisture and emissivity (cloud masked and atmospherically corrected) than with normalized brightness temperature. Table 3 shows that this is not the case. Although the correlation coefficient is higher for May and June, there are fewer data points and the relationship is less significant.

[28] A comparison between the slope values for the satellite data in Table 3 and those for the aircraft data in Table 2 shows that the satellite measurements have about half the sensitivity to soil moisture as the aircraft bare soil observations.

[29] For applications that require surface emissivity information, it is interesting to study the variability of the emission. Table 4 summarizes the mean emissivity and standard deviation for each month. The mean monthly emissivity for a given channel does not change significantly over the time period despite the variations in soil moisture (Figure 2) and vegetation cover (Figure 3). The standard deviation is highest for the horizontal channels and for the lower frequency (19 GHz), which is to be expected since these channels are most sensitive to soil moisture variations.

Table 4. Mean SSM/I 19 GHz H Emissivity and Standard Deviation by Month
MonthFrequency, GHzMean H EmissivityStandard Deviation in H EmissivityMean V EmissivityStandard Deviation in V Emissivity
May190.8920.0460.9390.028
May370.9100.0260.9430.016
June190.8980.0310.9430.017
June370.9070.0220.9410.013
July190.9040.0340.9380.017
July370.9110.0160.9340.008

8. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[30] The spatial variability of soil moisture is of concern for the validation of remotely sensed observations. Average soil moisture measured by TDR probes at four sites agreed to within 4% soil moisture by volume with the average value of at least fifty gravimetric measurements distributed throughout the agricultural study area. This indicates that in an area of flat topography and fairly uniform soil texture it may be possible to capture changes in areally averaged soil moisture with a limited amount of measurements.

[31] A very weak relationship was found between soil moisture and 19 GHz H microwave emission measured by aircraft or satellite. The standard error in the soil moisture estimate from the aircraft observations was 7% by volume or greater as opposed to 5% by volume or less for a grassland region in Oklahoma [Jackson, 1997]. The relatively high vegetation cover and variability in crop types in this type of agricultural area makes it difficult to obtain useful information on soil moisture from SSM/I data at the start of the growing season and impossible in the summer months. Interestingly, there was a weaker relationship between soil moisture and emissivity than between soil moisture and brightness temperature. This may be due to lack of accurate atmospheric information for the emissivity calculation. However, the same result was obtained with the aircraft data.

[32] The calculation of microwave emissivity for applications such as atmospheric sounding requires accurate surface temperature estimates. There was good agreement (within 1 K) between average surface temperature calculated from GOES-8 infrared data and in situ measurements made just above the surface at an automatic weather station.

[33] Despite changes in soil moisture and vegetation cover, the mean monthly emissivity in the study area did not change significantly over the May to July period. From the point of view of atmospheric sounding applications, this is a useful result since it suggests that spatial and temporal variations in emissivity are smoothed out at high frequencies.

[34] Lower-frequency radiometers operating at 1.4 and 6.9 GHz are currently being installed and tested on the Twin Otter aircraft. Future field experiments will test the ability of these sensors to retrieve soil moisture using data acquired over different Canadian land cover types.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information

[35] The aircraft measurements were made by the staff of the NRC and MSC Aircraft Facility. Dave McNichol, Arvids Silis and Eddie Graham worked hard to collect the ground truth data. We are grateful for the help of the Manitoba and Canadian Remote Sensing Centres in supplying NDVI data.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information
  • Bennartz, R., K. Paape, J. Fischer, and T. J. Hewison, Comparison of observed and simulated microwave land surface emissivities over bare soil, Meteorol. Z., 11, 512, 2002.
  • Duke, C., L. R. Protz, G. Parkin, P. von Bertoldi, A. J. VandenBygaart, Soil moisture estimation using 1. 4 GHz passive microwave radiometer data from local to regional scales, paper presented at 14th Conference on Hydrology, The 79th Annual Meeting of The American Meteorological Society,, Am. Meteorol. Soc., Dallas, Tex., 10 – 15 Jan. 1999.
  • Goita, K., A. Royer, and N. Bussieres, Analysis of land surface temperature and emissivities over a northern environment, paper presented at the 26th International Symposium on Remote Sensing of the Environment, Can. Aeronaut. and Space Inst.Vancouver, Canada, 25 – 29 March 1996.
  • Hollinger, J. P., J. L. Pierce, and G. A. Poe, SSM/I instrument evaluation, IEEE Trans, Geosci. Remote Sens., 28,5, 781790, 1990.
  • Hook, W. R., and N. J. Livingston, Errors in converting time domain reflectrometry measurements of propagation velocity to estimates of soil water content, Soil Sci. Soc. Am. J., 60, 3541, 1995.
  • Jackson, T. J., Soil moisture estimation using special satellite microwave/imager satellite data over a grassland region, Water Resour. Res., 33(6), 14751484, 1997.
  • Jones, A. S., and T. H. Vonder Haar, Retrieval of microwave surface emittance over land using coincident microwave and infrared satellite measurements, J. Geophys. Res., 102(D12), 13,60913,626, 1997.
  • Liebe, H. J., MPM—An atmospheric millimeter-wave propagation model, Int. J. Infrared Millimeter Waves, 10(6), 631650, 1989.
  • Liebe, H. J., G. A. Hufford, and M. G. Cotton, Propagation modeling of moist air and suspended water/ice particles at frequencies below 1000 GHz, paper presented at 52nd Specialists Meeting of the Electromagnetic Wave Propagation Panel, Advisory Group for Aerospace Res. and Dev., Palma de Mallorca, Spain, 1993.
  • MacPherson, J. I., D. L. Marcotte, and J. E. Jordan, The NRC Atmospheric Research Aircraft, Can. Aeronaut. Space J., 47(3), 111, 2001.
  • Morland, J., D. Grimes, G. Dugdale, and T. Hewison, The estimation of land surface emissivities at 24 to 157 GHz using remotely sensed aircraft data, Remote Sens. Environ., 73, 323336, 2000.
  • Morland, J. C., D. I. F. Grimes, and T. Hewison, Satellite observations of the microwave emissivity of a semi-arid land surface, Remote Sens. Environ., 77, 149164, 2001.
  • Petty, G. W., and K. B. Katsaros, The response of the SSM/I to the marine environment. part II: A parameterisation of the effect of the sea surface slope distribution on emission and reflection, J. Atmos. Oceanic Technol., 11(3), 617628, 1994.
  • Prigent, C., W. B. Rossow, and E. Matthews, Microwave land surface emissivities estimated from SSM/I observations, J. Geophys. Res., 102(D18), 21,86721,890, 1997.
  • Vinnikov, K. Y., A. Robock, S. Qui, J. K. Entin, M. Owe, B. J. Choudhury, S. E. Hollinger, and E. G. Njoku, Satellite remote sensing of soil moisture in Illinois, United States, J. Geophys. Res., 104(D4), 41454168, 1999.
  • Wang, J., Microwave emission from smooth bare fields and soil moisture sampling depth, IEEE Trans. Geosci. Remote Sens., 25, 616622, 1987.
  • Wilheit, T., Radiative transfer in a plane stratified dielectric, IEEE Trans. Geosci. Electron., 16, 138143, 1978.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Field Campaign and Aircraft Data
  5. 3. Satellite Data
  6. 4. Calculation of Emissivity
  7. 5. Aircraft Analysis
  8. 6. Comparison of Aircraft and Satellite Data
  9. 7. Satellite Analysis
  10. 8. Conclusions
  11. Acknowledgments
  12. References
  13. Supporting Information
FilenameFormatSizeDescription
rds4870-sup-0001tab01.txtplain text document0KTab-delimited Table 1.
rds4870-sup-0002tab02.txtplain text document0KTab-delimited Table 2.
rds4870-sup-0003tab03.txtplain text document1KTab-delimited Table 3.
rds4870-sup-0004tab04.txtplain text document0KTab-delimited Table 4.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.