2. COSMIC temperature and water vapor retrievals

COSMIC observations distribute relatively uniformly in space and time (see Figure 1). By placing a GPS receiver in LEO, GPS RO technique measures the phase delay of the radio signal from the GPS constellation precisely as the signal traverses the Earth's atmosphere. Being an active limb-sounding measurement, GPS RO technique is capable of retrieving profiles of microwave refractivity at very high vertical resolution [23]. The root mean square (RMS) error was estimated to be less than 1 K based on a detailed theoretical study [23], and this estimate was found to be consistent with numerous cross-validation studies between RO, radiosonde observation (RAOB), and other satellite measurements (e.g., [20, 21, 24–28]).

Although RO measurements are not sensitive to clouds, they are very sensitive to the vertical structure of atmospheric density profiles (a function of temperature, pressure, and water vapor profiles). When accurate RO observations, precise positions, and velocities of GPS and LEO satellites are known, accurate atmospheric temperature and moisture profiles can be derived [15, 29]. In a neutral atmosphere,

#### Figure 1.

Typical distribution of COSMIC GPS radio occultation soundings (green dots) over a 24-h period over the global. Red dots are the distribution of operational radiosonde stations.

and climate studies in terms of vertical resolution, spatial and temporal coverage, and accuracy. Satellite infrared (IR) and microwave sounders have been routinely used for monitoring the temperature and moisture profiles in the mid and lower troposphere since 1980. Launched in 2002, Atmospheric InfraRed Sounder (AIRS) is a high spectral resolution radiometer onboard NASA Aqua satellite. With those more than 2000 high spectrum resolution channels in infrared wavebands, AIRS is able to provide excellent temperature and water vapor retrievals at the midtroposphere level under clear skies. The AIRS measurements, together with more recent high spectrum resolution infrared (IR) measurements from Atmospheric Sounding Interferometer (IASI, 2006–current), and Cross-track Infrared Sounder (CrIS, 2011–current) have maintained continuous observations of tropospheric water vapor since 2002. However, due to the limitation of resolving power in terms of weighting functions and signal to noise ratio of these instruments, accurate estimates of water vapor concentration in the lower troposphere (LT) are still not available. Infrared sounders cannot sense atmospheric profiles below clouds. While infrared data are limited to clear skies, microwave (MV) sounders can provide all sky data products. There are three main microwave radiometers with sufficient resolution and stability to measure tropospheric water vapor: the Advanced Microwave Sounding Unit-A (AMSU-A) and Advanced Microwave Sounding Unit-B (AMSU-B) on NOAA-15, 16, and 17 (i.e., N15, N16, and N17) satellites, the Microwave Humidity Sounder (MHS) on NOAA-18 (N18) and MetOp-A (Meteorological Operational satellite A) satellites, and Advanced Technology Microwave Sounder (ATMS) on Suomi National Polar orbiting Partnership (SNPP) and the first Joint Polar Satellite System (JPSS-1). These microwave

sounders onboard the polar-orbiting satellites have been routinely used by NOAA to generate the tropospheric temperature and moisture profiles for all-sky conditions and hydrological variables (e.g., rainfall, precipitable water, cloud water, ice water,

Using a one-dimensional variational (1DVAR) scheme, the NOAA Microwave Integrated Retrieval System (MiRS) inversion package is routinely applied to the AMSU/MHS sensors onboard the NOAA-18 (N18), NOAA-19 (N19), and Meteorological Operational Satellite-A (Metop-A) satellites to optimally retrieve temperature, moisture, and surface skin temperature (SST), as well as hydrometer variables within clouds over land and oceans [9, 10]. The MiRS-retrieved parameters have been validated globally using independent measurements [9, 11, 12]. However, studies demonstrated that the MiRS-derived hydrometer parameters within clouds over lands and oceans still contain uncertainty, especially in the lower troposphere [10]. This is partly because there is not enough information from the AMSU/ATMS measurements to completely resolve the hydrometer variables, temperature, and

In addition to the passive infrared and microwave sounder observations, the active global positioning system (GPS) radio occultation (RO) technique can also provide all-weather temperature and moisture profiles. Unlike passive MW and IR sensors, GPS RO is an active remote sensing technique that can provide all-weather, high vertical resolution (from 100 m near the surface to 1.5 km at 40 km) bending angle, and refractivity profiles [13, 14]. With knowledge of the precise positions and velocities of the GPS and low earth orbiting (LEO) satellites, which carry the GPS receivers, a vertical distribution of bending angles at the ray perigee point (the point of the ray path that is closest to Earth) can be derived. From the vertical distribution of the bending angle, we can derive a vertical distribution of the atmospheric bending angle and refractivity, which is a function of atmospheric temperature, moisture, and pressure [13, 14]. The GPS RO data have been intensively used for weather prediction and climate studies since the launch of the

etc.) under cloudy conditions.

Green Chemistry Applications

water vapor profiles under clouds.

90

the refractivity (N) is related to the pressure (P), the temperature (T), and the partial pressure of WV (PW) as represented by the following equation:

$$N = 77.6 \frac{P}{T} + 3.73 \times 10^5 \frac{P\_W}{T^2} \tag{1}$$

and Metop-A CDAAC products, Figure 2 shows a statistical comparison for July 2014 of co-located COSMIC postprocessed (consistent with COSMIC2013) and Metop-A2016 reprocessed bending angle profiles. The match criteria used were: time differences <90 min and distances <250 km. The mean differences are very small. The differences in standard deviation at 30 km altitude (south pole region larger than north pole region) are believed to be due to stronger horizontal variability of the atmosphere during local winter. Remaining GPS RO missions will be processed as soon as possible, and all data will be made available via

Global Water Vapor Estimates from Measurements from Active GPS RO Sensors and Passive…

RO derived temperature profiles especially in the lower stratosphere have also been intensively validated. Figure 3 depicts that COSMIC temperature is very close to those from Vaisala-RS92 from 200 to 20 hPa (around 12 to 25 km). Note that, Vaisala-RS92 is one of the most accurate modern radiosondes where the structural uncertainties are 0.2 K below 100 hPa and somewhat higher at higher levels. According to RS92 data continuity link under the Vaisala website, the Vaisala data including RS92 have been corrected for possible radiation errors. Their mean temperature difference in this height range is very close to zero. Because the quality of RO data does not vary with geophysical location and time, it is very useful to assess systematic errors of radiosonde sensors, whose characteristics may be affected by the changing environment and sensor types (e.g., [26, 28]). This result also shows

the quality of RO temperature profiles where the precision of the mean of COSMIC-derived temperature profiles is estimated to be better than 0.05 K from

Statistical comparison of co-located COSMIC postprocessed and Metop-A reprocessed bending angle profiles (red = mean and blue = standard deviation). Match criteria: time differences <90 min and distances <250 km.

CDAAC's website.

DOI: http://dx.doi.org/10.5772/intechopen.79541

8 to 30 km [18].

Figure 2.

93

Above the UT (�8 km) where moisture is negligible, the dry temperature and the actual temperature are nearly equal [23]. The accuracy, precision, and long-term stability of RO data have been quantified by many studies under various atmospheric conditions [18–21, 26–28, 30]. GPS RO measurements have many important attributes that make them uniquely suitable for climate monitoring. These include: (i) no satellite-to-satellite bias, (ii) no instrument drift, and (iii) not affected by clouds and precipitation.

#### 2.1 COSMIC temperature uncertainty

The distribution of water vapor profile depends on temperature [31, 32]. COS-MIC Data Analysis and Archive Center (CDAAC) is an international operational center that routinely inverses RO excess phase data obtained from multiple RO mission to bending angle and refractivity profiles. Recently, CDAAC has developed and installed new and improved RO inversion software, including improvements to precise orbit determination (POD), excess atmospheric phase computation, and neutral atmospheric inversion processing. Now these consistent RO inversion algorithms are applied to several international RO missions to derive the vertical distribution of bending angle, refractivity, temperature, and geo-potential height. These RO missions include GPS/MET (from 1995 to 1997, no overlap with other RO missions), COSMIC (launched in April 2006), Challenging Mini-satellite Payload (CHAMP, from 2001 to 2008), Gravity Recovery And Climate Experiment (GRACE, launched in 2004), Satélite de Aplicaciones Científicas-C (SAC-C, launched in 2000), GNSS RO Receiver for Atmospheric Sounding (GRAS, launched in 2007), Communication/Navigation Outage Forecast System (C/NOFS, launched in 2008), and Terra Synthetic Aperture Radar (SAR) operating in the X-band (TerraSAR-X, launched in 2007).

Currently, multi-year GPS RO climate data can be obtained from the GeoForschungsZentrum Potsdam (GFZ), Germany, the Jet Propulsion Laboratory (JPL), Pasadena, CA, USA, the University Corporation for Atmospheric Research (UCAR), Boulder, CO, USA, and the Wegener Center of the University of Graz (WegC), Graz, Austria. Different centers have used different assumptions, initializations, and implementations in the excess phase processing and inversion procedures, which may introduce refractivity differences between centers. Ho et al. [17] have used 5 years (2002–2006) of monthly mean climatologies (MMC) of retrieved refractivity from CHAMP generated by the above four centers to quantify the processing procedure-dependent errors. Results show that the absolute values of fractional refractivity anomalies among the centers are in general ≤0.2% from 8 to 25 km altitude (not shown). The median absolute deviation among the centers is less than 0.2% globally. This provides useful bounds on the errors introduced by data processing schemes.

The near real-time, postprocessing, and reprocessing status for all RO missions at CDAAC are summarized on the CDAAC website (http://cdaac-www.cosmic.uca r.edu/cdaac/products.html). A new version of Metop-A reprocessing, named Metop-A2016, has just finished processing and validation and will be posted to the CDAAC website in the next month. To illustrate the consistency between COSMIC Global Water Vapor Estimates from Measurements from Active GPS RO Sensors and Passive… DOI: http://dx.doi.org/10.5772/intechopen.79541

and Metop-A CDAAC products, Figure 2 shows a statistical comparison for July 2014 of co-located COSMIC postprocessed (consistent with COSMIC2013) and Metop-A2016 reprocessed bending angle profiles. The match criteria used were: time differences <90 min and distances <250 km. The mean differences are very small. The differences in standard deviation at 30 km altitude (south pole region larger than north pole region) are believed to be due to stronger horizontal variability of the atmosphere during local winter. Remaining GPS RO missions will be processed as soon as possible, and all data will be made available via CDAAC's website.

RO derived temperature profiles especially in the lower stratosphere have also been intensively validated. Figure 3 depicts that COSMIC temperature is very close to those from Vaisala-RS92 from 200 to 20 hPa (around 12 to 25 km). Note that, Vaisala-RS92 is one of the most accurate modern radiosondes where the structural uncertainties are 0.2 K below 100 hPa and somewhat higher at higher levels. According to RS92 data continuity link under the Vaisala website, the Vaisala data including RS92 have been corrected for possible radiation errors. Their mean temperature difference in this height range is very close to zero. Because the quality of RO data does not vary with geophysical location and time, it is very useful to assess systematic errors of radiosonde sensors, whose characteristics may be affected by the changing environment and sensor types (e.g., [26, 28]). This result also shows the quality of RO temperature profiles where the precision of the mean of COSMIC-derived temperature profiles is estimated to be better than 0.05 K from 8 to 30 km [18].

#### Figure 2.

Statistical comparison of co-located COSMIC postprocessed and Metop-A reprocessed bending angle profiles (red = mean and blue = standard deviation). Match criteria: time differences <90 min and distances <250 km.

the refractivity (N) is related to the pressure (P), the temperature (T), and the

Above the UT (�8 km) where moisture is negligible, the dry temperature and the actual temperature are nearly equal [23]. The accuracy, precision, and long-term

The distribution of water vapor profile depends on temperature [31, 32]. COS-MIC Data Analysis and Archive Center (CDAAC) is an international operational center that routinely inverses RO excess phase data obtained from multiple RO mission to bending angle and refractivity profiles. Recently, CDAAC has developed and installed new and improved RO inversion software, including improvements to precise orbit determination (POD), excess atmospheric phase computation, and neutral atmospheric inversion processing. Now these consistent RO inversion algorithms are applied to several international RO missions to derive the vertical distribution of bending angle, refractivity, temperature, and geo-potential height. These RO missions include GPS/MET (from 1995 to 1997, no overlap with other RO missions), COSMIC (launched in April 2006), Challenging Mini-satellite Payload (CHAMP, from 2001 to 2008), Gravity Recovery And Climate Experiment (GRACE, launched in 2004), Satélite de Aplicaciones Científicas-C (SAC-C, launched in 2000), GNSS RO Receiver for Atmospheric Sounding (GRAS, launched in 2007), Communication/Navigation Outage Forecast System (C/NOFS, launched in 2008), and Terra Synthetic Aperture Radar (SAR) operating in the X-band

Currently, multi-year GPS RO climate data can be obtained from the GeoForschungsZentrum Potsdam (GFZ), Germany, the Jet Propulsion Laboratory (JPL), Pasadena, CA, USA, the University Corporation for Atmospheric Research (UCAR), Boulder, CO, USA, and the Wegener Center of the University of Graz (WegC), Graz, Austria. Different centers have used different assumptions, initializations, and implementations in the excess phase processing and inversion procedures, which may introduce refractivity differences between centers. Ho et al. [17] have used 5 years (2002–2006) of monthly mean climatologies (MMC) of retrieved refractivity from CHAMP generated by the above four centers to quantify the processing procedure-dependent errors. Results show that the absolute values of fractional refractivity anomalies among the centers are in general ≤0.2% from 8 to 25 km altitude (not shown). The median absolute deviation among the centers is less than 0.2% globally. This provides useful bounds on the errors introduced by

The near real-time, postprocessing, and reprocessing status for all RO missions at CDAAC are summarized on the CDAAC website (http://cdaac-www.cosmic.uca r.edu/cdaac/products.html). A new version of Metop-A reprocessing, named Metop-A2016, has just finished processing and validation and will be posted to the CDAAC website in the next month. To illustrate the consistency between COSMIC

<sup>T</sup> <sup>þ</sup> <sup>3</sup>:<sup>73</sup> � <sup>10</sup><sup>5</sup> PW

<sup>T</sup><sup>2</sup> (1)

partial pressure of WV (PW) as represented by the following equation:

stability of RO data have been quantified by many studies under various atmospheric conditions [18–21, 26–28, 30]. GPS RO measurements have many important attributes that make them uniquely suitable for climate monitoring. These include: (i) no satellite-to-satellite bias, (ii) no instrument drift, and (iii) not

affected by clouds and precipitation.

Green Chemistry Applications

(TerraSAR-X, launched in 2007).

data processing schemes.

92

2.1 COSMIC temperature uncertainty

<sup>N</sup> <sup>¼</sup> <sup>77</sup>:<sup>6</sup> <sup>P</sup>

#### Figure 3.

Comparisons of temperature between COSMIC and radiosonde for Vaisala-RS92. Mean bias, absolute mean bias, and mean standard deviation from 200 to 10 hPa are computed. The red line is the mean difference, the blue line is the standard deviation, and the dotted line is the sample number. The top X axis shows the sample number.

COSMIC data have been used to study atmospheric temperature and refractivity trends in the lower stratosphere [17–19, 22] and variation of water vapor above, within, and below clouds [36–43]. COSMIC water vapor data have also been used to detect climate signals like El Niño-Southern Oscillation (ENSO; [38–40]), Madden-Julian Oscillation (MJO; [41]), atmospheric rivers [42, 43], and TPW variation

Left panel: the global comparisons of total precipitable water (TPW) between COSMIC and those derived from ground-based GPS (i.e., IGS) for 2008. The right panel indicates the time series of IGS TPW and COSMIC-

Global Water Vapor Estimates from Measurements from Active GPS RO Sensors and Passive…

In this study, we will use COSMIC data collocated with AIRS (COSMIC-AIRS pairs) and ATMS (COSMIC-ATMS pairs) to derive the temperature and moisture profiles. The relatively uniformly distributed COSMIC profiles in space and time would allow numerous RO and AIRS/ATMS coincident pairs, which would provide unprecedented atmospheric temperature and moisture profiles under various

As mentioned above, MiRS is a MW inversion package used by NOAA to perform a physical-based microwave retrieval in all-weather scenarios over all landsurface types. The MiRS data products have been routinely assessed using independent measurements [9, 11, 12]. The MiRS package can be downloaded, for free, at http://mirs.nesdis.noaa.gov/download.php. In this study, we revised the current MiRS algorithm and use measurements from ATMS collocate with RO data to develop an enhanced RO-ATMS inversion algorithm (i.e., RO-MiRS). We included the RO refractivity forward operator (Eq. (1)) into the RO-MiRS algorithm.

Ref. [34] has detailed the information content for AIRS, RO, and the combined AIRS with RO for temperature and water vapor retrievals. Similar to the AIRS V6, the RO-AIRS inversion system is a 1D-var physical inversion system that retrieves the temperature and water vapor profiles sequentially. The updated fast and accurate AIRS transmittance model (Standard Alone AIRS-Radiance Transfer Algorithm

owing to global warming [44–46].

DOI: http://dx.doi.org/10.5772/intechopen.79541

derived TPW near the IGS station.

Figure 4.

3.1 RO-ATM inversion algorithm

3.2 RO-AIRS inversion algorithm

95

atmospheric conditions, which was not possible before.

3. The RO-ATMS and RO-AIRS inversion algorithms

#### 2.2 COSMIC water vapor uncertainty

By using an advanced tracking technique, known as "open-loop tracking" [33], more than 90% of RO profiles from the COSMIC mission penetrate to below 2 km. As shown in Eq. (1), GPS RO refractivity is sensitive to temperature or water vapor, depending on the atmospheric conditions. Ho et al. [34] showed that in the upper troposphere where water vapor is negligible, RO observations are more sensitive to atmospheric temperature variations than to water vapor content. However, in the moisture rich troposphere, the RO refractivity is more sensitive to water vapor variation [34].

A 1D-var algorithm (http://cosmic-io.cosmic.ucar.edu/cdaac/doc/documents/1d var.pdf) is used to derive optimal temperature and water vapor profiles while temperatures and water vapor profiles from RO refractivity (see Eq. (1)). The ERA-Interim reanalyzes are used as a priori estimates for the 1D-var algorithm. The accuracy of COSMIC-derived total precipitation water (TPW) has been demonstrated by comparisons with TPW derived from ground-based GPS (i.e., International Global Navigation Satellite Systems (IGS, [20, 35])) which are assumed not to be geolocation dependent. Figure 4 (left panel) depicts COSMIC TPW and those from ground-based GPS collected within 2 h and 200 km in 2008. Only those COSMIC profiles whose lowest penetration heights are within 200 m of the height of ground-based GPS stations are included. Results demonstrate that the mean TPW difference between IGS and COSMIC is less than 0.2 mm with a standard deviation of 2.69 mm. This demonstrates the accuracy of COSMIC-derived water vapor in the lower troposphere, which should be particularly useful for improving AIRS/ATMS retrievals, especially over ice and cold surface backgrounds (see Sections 4 and 5). The right panel in Figure 4 shows the time series of the COSMIC TPW and those from IGS at the same station. This demonstrates the importance and usefulness of COSMIC RO observations in depicting global water vapor variations.

Global Water Vapor Estimates from Measurements from Active GPS RO Sensors and Passive… DOI: http://dx.doi.org/10.5772/intechopen.79541

Figure 4.

2.2 COSMIC water vapor uncertainty

variation [34].

94

Figure 3.

Green Chemistry Applications

number.

By using an advanced tracking technique, known as "open-loop tracking" [33], more than 90% of RO profiles from the COSMIC mission penetrate to below 2 km. As shown in Eq. (1), GPS RO refractivity is sensitive to temperature or water vapor, depending on the atmospheric conditions. Ho et al. [34] showed that in the upper troposphere where water vapor is negligible, RO observations are more sensitive to atmospheric temperature variations than to water vapor content. However, in the moisture rich troposphere, the RO refractivity is more sensitive to water vapor

Comparisons of temperature between COSMIC and radiosonde for Vaisala-RS92. Mean bias, absolute mean bias, and mean standard deviation from 200 to 10 hPa are computed. The red line is the mean difference, the blue line is the standard deviation, and the dotted line is the sample number. The top X axis shows the sample

A 1D-var algorithm (http://cosmic-io.cosmic.ucar.edu/cdaac/doc/documents/1d

var.pdf) is used to derive optimal temperature and water vapor profiles while temperatures and water vapor profiles from RO refractivity (see Eq. (1)). The ERA-Interim reanalyzes are used as a priori estimates for the 1D-var algorithm. The accuracy of COSMIC-derived total precipitation water (TPW) has been demonstrated by comparisons with TPW derived from ground-based GPS (i.e., International Global Navigation Satellite Systems (IGS, [20, 35])) which are assumed not to be geolocation dependent. Figure 4 (left panel) depicts COSMIC TPW and those from ground-based GPS collected within 2 h and 200 km in 2008. Only those COSMIC profiles whose lowest penetration heights are within 200 m of the height of ground-based GPS stations are included. Results demonstrate that the mean TPW difference between IGS and COSMIC is less than 0.2 mm with a standard deviation of 2.69 mm. This demonstrates the accuracy of COSMIC-derived water vapor in the lower troposphere, which should be particularly useful for improving AIRS/ATMS retrievals, especially over ice and cold surface backgrounds (see Sections 4 and 5). The right panel in Figure 4 shows the time series of the COSMIC TPW and those from IGS at the same station. This demonstrates the importance and usefulness of

COSMIC RO observations in depicting global water vapor variations.

Left panel: the global comparisons of total precipitable water (TPW) between COSMIC and those derived from ground-based GPS (i.e., IGS) for 2008. The right panel indicates the time series of IGS TPW and COSMICderived TPW near the IGS station.

COSMIC data have been used to study atmospheric temperature and refractivity trends in the lower stratosphere [17–19, 22] and variation of water vapor above, within, and below clouds [36–43]. COSMIC water vapor data have also been used to detect climate signals like El Niño-Southern Oscillation (ENSO; [38–40]), Madden-Julian Oscillation (MJO; [41]), atmospheric rivers [42, 43], and TPW variation owing to global warming [44–46].

In this study, we will use COSMIC data collocated with AIRS (COSMIC-AIRS pairs) and ATMS (COSMIC-ATMS pairs) to derive the temperature and moisture profiles. The relatively uniformly distributed COSMIC profiles in space and time would allow numerous RO and AIRS/ATMS coincident pairs, which would provide unprecedented atmospheric temperature and moisture profiles under various atmospheric conditions, which was not possible before.
