Cite this Article
CHEN Ting, WANG Dong-Hai, LI Guo-Ping, ZHANG Yu-Wei, 2013: Validation of IASI-Retrieved Atmospheric Water Vapor Data over the Tibetan Plateau Region. Atmos Oceanic Sci Lett, 6(3): 167-172  
Permissions
Copyright©2013, Institute of Atmospheric Physics, Chinese Academy of Sciences
Validation of IASI-Retrieved Atmospheric Water Vapor Data over the Tibetan Plateau Region
CHEN Ting1,2, WANG Dong-Hai2*, LI Guo-Ping1, ZHANG Yu-Wei2
1College of Atmospheric Sciences, Chengdu University of Information Technology, Chengdu 610225, China 2State Key Laboratory of Severe Weather, Chinese Academy of Meteorological Sciences, Beijing 100081, China
* Corresponding author: WANG Dong-Hai, E-mail:wangdh@cams.cma.gov.cn
Fund:Science and Technology (Grant Nos. GYHY200806007, GYHY 201006014, and GYHY201206039), and the National Natural Science Foundation (Grant Nos. 41175064, 40875022, and 40633016). We are grateful to two anonymous reviewers for their valuable comments and suggestions to improve the quality of the manuscript.
Abstract

The Infrared Atmospheric Sounding Interferometer (IASI) is a new-generation ultraspectral atmospheric sounding instrument mounted on the MetOp-A, the first operational polar-orbiting satellite developed by the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT). It is an ultrahigh spectral-resolution atmospheric detector which can detect atmospheric chemical composition, temperature, and humidity profiles with high accuracy and resolution. In the present study, through comparative analyses of the similarities and differences between the IASI and the radiosonde observation (RAOB) water vapor data, and between the IASI and the Aqua-AIRS water vapor retrievals, a detailed and systematic assessment of the credibility of the IASI water vapor retrievals over the plateau region was made. A comparison of the IASI retrievals with the AIRS retrievals and the RAOB measurements over the Tibetan Plateau revealed that the IASI retrieval data are reliable and can be used for conducting further studies.

Key words: IASI; Tibetan Plateau; water vapor profiles; quality validation
1 Introduction

The Tibetan Plateau is considered as an indicator of as well as a contributor to the climatic changes occurring in the Northern Hemisphere ( Feng et al. , 1998; Liu and Chen, 2000); it also plays an important role in severe weather events ( Xu, 2009). The vortexes that are generated over the plateau and move eastward act as a trigger for the southwest vortex and severe rainfall around the Tibetan Plateau ( Li, 2007). The Tibetan Plateau influences not only the weather and the climate of China, but also those around the world ( Ye et al. , 1957; Wu et al. , 2005). Since the Tibetan Plateau acts as a heat pump in summer, and the water vapor from the Bay of Bengal or Arabian Sea converges on the Tibetan Plateau, the plateau becomes a transferring station for water vapor. On the one hand, water vapor is transported to the east of the Tibetan Plateau, causing severe weather events ( Xu et al. , 2001, 2002), and on the other hand, it is transported upward to the stratosphere ( Gettelman et al. , 2004; Fu et al. , 2006). Therefore, it is very important to know the distribution of the water vapor over the Tibetan Plateau. Although the radiosonde observation (RAOB) water vapor data can be obtained easily, these are not sufficient because of the limited temporal and spatial distribution of the stations, especially over the Tibetan Plateau, due to its complex terrain.

In the recent years, ground-based remote sensing ( Li and Huang, 2004) and satellite remote sensing methods ( Li and Zhu, 1998) are also used to derive the water vapor profile. The infrared atmospheric sounding satellite is mainly used for remote sensing of atmospheric temperature, humidity profiles, and atmospheric composition. A filter infrared atmospheric sounding instrument such as the High-Resolution Infrared Radiation Sounder (HIRS) was developed by the United States in the 1970s, but limited by the technologies, its resolution could hardly be enhanced. Consequently, scientists began to shift their emphasis on the development of high-spectral-resolution atmospheric detectors, such as the Atmospheric Infrared Sounder (AIRS) ( Zhang et al. , 2012, 2013), and infrared atmospheric sounder interferometers, such as the Infrared Atmospheric Sounding Interferometer (IASI), a new-generation ultraspectral atmospheric sounding instrument. The IASI is mounted on the MetOp-A satellite, and its main goal is to provide atmospheric emission spectra from which temperature and humidity profiles can be derived with high vertical resolution and accuracy. Pougatchev et al. (2009) reported that the IASI water vapor retrievals were in good agreement with the RAOB data in the range from 800 to 300 hPa . However, research on the validation of the IASI water vapor retrievals over the Tibetan Plateau is still at an experimental stage. In this paper, the IASI water vapor retrievals were compared with those of the RAOB and the AIRS retrievals to help validate the suitability of the IASI water vapor retrievals over the Tibetan Plateau and provide a reference for relevant water vapor studies.

2 Data and methods
2.1 Data

Three following data sets, covering a period of four years (2008-2011), were used for the study:(1) The water vapor profiles of level 2 products of the IASI mounted on the MetOp-A satellite were downloaded from the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) server. The vertical resolution of the IASI humidity data was 1-2 km in the lower troposphere and the horizontal resolution was 25 km, with an accuracy of 10%. The water vapor data had 90 standard layers from 1000 to 100 hPa. In order to compare these data with the 12 standard layers of daily data of the RAOB and the 12 standard layers of monthly data of AIRS, they were interpolated using the Inverse Distance Weighted (IDW) technique to 12 standard layersand processed as daily and monthly data. IASI data were interpolated to the stations when comparing with the RAOB data and to the grids when comparing with the AIRS data.

(2) The temperature and dew point temperature data of the RAOB stations over the Tibetan Plateau were collected from the China National Meteorological Information Center. After being revised, the data passed the homogeneity test. In this study, 18 RAOB stations over the Tibetan Plateau, situated between 25-40°N and 75-100°E at an altitude of higher than 2000 m above sea level, were selected; vertical profiles of the temperature and dew point temperature had 12 standard layers ranging from 1000 to 100 hPa, at a frequency of two observations per day. They were processed as daily data. It is worth mentioning here that a lot of dew point temperature data above 200 hPa were missing, and as the altitude of the Tibetan Plateau is very high, 700 hPa was considered to be the lower level of the plateau. Figure 1 shows the spatial distribution of the RAOB stations and the altitude of the terrain. These stations cover most of the places in the Tibetan Plateau except its western part.

(3) The data on water vapor mixing ratio over the Tibetan Plateau were retrieved from the AIRS mounted on the National Aeronautics and Space Administration (NASA)/Aqua satellite. In this paper, Level 3 data were used, which were monthly data (for the period 2008-2011) with 12 standard layers ranging from 1000 to 100 hPa and a horizontal resolution of 1°×1°. It should be noted here that the water vapor mixing ratio of the lower layer was equal to the average of that of the adjacent two layers; for example, the water vapor mixing ratio of the layer at 200 hPa was actually the average of those of the layers at 200 and 150 hPa.

Fig. 1 Spatial distribution of the RAOB stations (black dots) in the Tibetan Plateau and the terrain (shaded, units: m).

2.2 Methods

In this study, the water vapor mixing ratio was computed based on the temperature and dew point temperature data. Water existed at T≥0°C and ice at T≤-20°C, with ice and water coexisting in the temperature range from 0 to -20°C ( Goff and Gratch, 1946). The IASI retrieval data were compared with the RAOB and the AIRS retrieval data to find out the differences among them. Then the RMS differences and the bias were computed to explain quantitatively the differences between the IASI water vapor data and the RAOB measurements and between the IASI and the AIRS retrievals in cold seasons (April-September) and warm seasons (October-March).

3 Results and discussion

The correlation coefficients and the RMS differences between IASI retrievals and the RAOB data, and those between IASI retrievals and the AIRS retrievals are discussed and analyzed in this section. The corresponding bias characteristics are discussed as well.

Figures 2a and 2b show the correlation coefficients between the IASI and the RAOB water vapor data over the Tibetan Plateau during the warm seasons (April- September) and the cold seasons (October-March), respectively, while Figs. 2c and 2d show the correlation coefficients between the IASI and the AIRS water vapor data during those two seasons. It is clear to see that both of the correlations are obvious, indicating that their consistencies are good. And the entire Plateau has passed the significant test of which significant level is 0.05. In both seasons, correlation between the IASI and the AIRS water vapor data was slightly better than that between the IASI and the RAOB data. On the other hand, the correlation coefficients were found to be higher in warm seasons than in cold seasons in both the cases.

Figure 3a shows the correlation coefficients between the IASI water vapor data and the RAOB data, and those between the IASI water vapor data and the AIRS retrievals in warm seasons over the Tibetan Plateau, while Fig. 3d shows the corresponding correlation coefficients in cold seasons. Due to the missing of the RAOB dew point data at 100 hPa, the correlation coefficients, RMS differences, and bias for this layer could not be calculated. In both cold and warm seasons, the correlation coefficients of the IASI water vapor data were in good agreement with those of the RAOB water vapor data between 700 and 250 hPa levels, with the best agreement being at the 500 hPa level. Both cases showed that the correlations began to decrease above 250 hPa, which might be attributed to the dramatic drifting of RAOB balloon and the missing dew point temperature data, both above 250 hPa. However, the correlation coefficients between the IASI and the AIRS water vapor data at the 12 standard levels were close to each other and slightly higher than those between the IASI and the RAOB data.

Figure 3b shows the vertical profiles of the RMS differences between the IASI water vapor data and the RAOB/the AIRS data in warm seasons over the Tibetan Plateau, while Fig. 3e shows the RMS differences in cold seasons. Irrespective of the seasons, the RMS differences at 500 and 600 hPa were lowest in both cases. However, the RMS differences of the warm seasons (within 30%) were smaller than those of the cold seasons (within 40%). The RMS differences between the IASI and the RAOB data were found to be large above 150 hPa, but the differences between the IASI and the AIRS data were small at all the levels.

Figures 3c and 3f show the characteristics of the water vapor bias profiles in warm seasons and cold seasons, respectively, as implied by the comparingson between the IASI and the RAOB and between the IASI and the AIRS retrieval data. From the comparison between the IASI and the RAOB data, it can be seen that the bias from the low level to 250 hPa was small in warm seasons, with the maximum of 25%. Negative water vapor bias existed between 500 and 400 hPa, which might be because the IASI retrievals having a dry bias in these layers, but having a wet bias in the other layers. In cold seasons, the IASI retrievals had a wet bias between 300 and 200 hPa but a dry bias between 700 and 400 hPa. On the other hand, the amplitude of bias between IASI and RAOB data was larger in cold seasons than in warm seasons, but it was still within the expected error bounds (25%). Comparing the IASI with the AIRS retrievals, the IASI water vapor were found to have a wet bias between 700 and 100 hPa both in warm and cold seasons. In warm seasons, the bias amplitude was 30% while in the cold seasons, it was 20%.

Figures 4a and 4b show the spatial distribution of the RMS differences between the IASI and the RAOB water vapor data over the Tibetan Plateau, while Figs. 4c and 4d show the RMS differences between the IASI and theAIRS retrievals in warm and cold seasons, respectively. With respect to the entire Tibetan Plateau region, the RMS differences between the IASI and the AIRS retrievals were slightly more than those between the IASI and the RAOB retrievals in both warm and cold seasons; the RMS differences in warm seasons were higher than those in cold seasons for both cases.

Figure 5 shows the spatial distribution of the water vapor bias over the Tibetan Plateau. Here, the RAOB measurements and the AIRS water vapor retrievals were considered as references. Differences between the IASI data and the references were calculated first, and then the bias was defined as the percentage of the difference between the IASI and the RAOB data (Figs. 5a and 5b) or between the IASI and the AIRS data (Figs. 5c and 5d); corresponding RMS differences for these biases are shown in Fig. 4. Biases in both cases were below 25%, except for a few stations. Biases of the cold seasons were smaller than those of the warm seasons. As far as the overall plateau is concerned, the bias between the IASI and the RAOB was smaller than that between the IASI and the AIRS. Some stations were found to have wet bias and some dry bias when the IASI and the RAOB data were compared, but almost all stations had wet bias when the IASI and the AIRS data were compared. Compared with the bias of the IASI retrievals at Lindenberg station ( Pougatchev et al. , 2009), the bias over the Tibetan Plateau was larger, which may be attributed to the complex terrain over the Tibetan Plateau and incorrect surface parameters.

Fig. 2 Spatial distribution of the correlation coefficients between the IASI and (a, b) the RAOB retrieval data and between the IASI and (c, d) the AIRS data over the Tibetan Plateau during 2008-2011. (a) and (c) for warm season; (b) and (d) for cold season.

4 Summary and conclusions

Comparisons of the IASI retrievals with the AIRS retrievals and the RAOB measurements over the Tibetan Plateau revealed that the IASI retrieval data are reliable and can be used to conduct further studies.

(1) The consistencies of the water vapor data retrievals from the IASI, with the RAOB and the AIRS retrievals are all high from 700 to 250 hPa. The difference between the IASI and the AIRS water vapor retrievals was higher than those between the IASI and the RAOB data.

(2) The RMS differences between the IASI and the AIRS retrievals were slightly better than those between the IASI and the RAOB measurements in both warm and cold seasons.

(3) With respect to the bias, the water vapor retrievals from the IASI matched both the RAOB water vapormeasurements (within 25%) and the AIRS retrievals (within 30%).

(4) The water vapor data retrievals from the IASI and the AIRS were in good agreement with each other, but larger differences were reported between the water vapor data from the IASI retrievals and the RAOB measurements in some areas above 250 hPa, which might be attributed to the upward drifting of the RAOB balloon.

(5) The IASI credibility in warm seasons was higher than that in cold seasons. Analyses of the correlation coefficients, RMS differences, and bias implied that the most credible level was 500 hPa.

(6) The IASI data were wetter than the RAOB data, except at 400 and 500 hPa in warm seasons and from 700 to 400 hPa in cold seasons, but wetter than AIRS data at all levels in both warm and cold seasons.

Fig. 3 Vertical profiles of the correlation coefficients, water vapor RMS differences, and water vapor biases between the IASI retrieval data and the RAOB observations/AIRS retrieval data over the Tibetan Plateau in (a-c) summer half year (April-September) and (d-f) winter half-year (October-March).
Fig. 4 Spatial distribution of the water vapor RMS differences between the IASI and (a, b) the RAOB retrieval data and between the IASI and (c, d) the AIRS data over the Tibetan Plateau during 2008-2011. (a) and (c) for warm season; (b) and (d) for cold season.
Fig. 5 Spatial distribution of the water vapor bias between the IASI and (a, b) the RAOB retrieval data and between the IASI and (c, d) the AIRS data over the Tibetan Plateau during 2008-2011. (a) and (c) for warm season; (b) and (d) for cold season.

References
1 Feng, S. , M. C. Tang, D. M. Wang, 1998: New evidence of Tibetan Plateau as the climate change start-up zone of China, Chinese Sci. Bull. (in Chinese), 43(6), 633-636.
2 Fu, R. , Y. Hu, J. S. Wright, et al. , 2006: Short circuit of water vapor and polluted air to the global stratosphere by convective transport over the Tibetan Plateau, Proc. Natl. Acad. Sci. USA, 103, 5664-5669.
3 Gettelman, A. , D. E. Kinnison, T. J. Dunkerton, et al. , 2004: The impact of monsoon circulations on the upper troposphere and lower stratosphere, J. Geophys. Res. , 109, doi: 10. 1029/2004 JD004878.
4 Goff, J. A. , S. Gratch, 1946: Low pressure properties of water from -160 F to 212F, Trans. Amer. Soc. Heat. Vent. Eng. , 52, 95-121.
5 Li, C. , Y. Zhu, 1998: The applied on retrievals of perceptible water from the GMS-5 'Split Window', Acta Sci. Nat. Univ. Peking (in Chinese), 34(1), 33-39.
6 Li, G. P. , 2007: Dynamic Meteorology of the Tibetan Plateau (2nd Ed. ), China Meteorological Press, Beijing, 20pp
7 Li, G. P. , D. F. Huang, 2004: Reviews and prospects of researches on remote sensing of regional atmospheric water vapor using ground-based GPS, Meteor. Sci. Technol. , 32(4), 201-205.
8 Liu, X. D. , B. D. Chen, 2000: Climatic warming in the Tibetan Plateau during recent decades, Int. J. Climate, 20(14), 1729-1742.
9 Pougatchev, N. , T. August, X. Calbet, et al. , 2009: IASI temperature and water vapor retrievals-Error assessment and validation, Atmos. Chem. Phys. , 9, 6453-6458.
10 Wu, G. X. , Y. M. Liu, X. Liu, et al. , 2005: How the heating over the Tibetan Plateau affects the Asian climate in summer, Chinese J. Atmos. Sci. (in Chinese), 29(1), 47-56.
11 Xu, X. D. , 2009: The effects of sensitive region over Tibetan Plateau on disastrous weather and climate and its monitoring, Eng. Sci. , 11(10), 96-107.
12 Xu, X. D. , M. Y. Zhou, J. Y. Chen, et al. , 2001: A comprehensive physical pattern of land-air dynamic and thermal structure on the Qinghai-Xizang Plateau, Sci. China Ser. D. -Earth Sci. (in Chinese), 31, 428-440.
13 Xu, X. D. , S. Y. Tao, J. Z. Wang, et al. , 2002: The relationship between water vapor transport features of Tibetan Plateau-monsoon 'Large Triangle' affecting region and drought-flood abnormality of ChinaActa Meteor. Sinica(in Chinese), 60, 257-266.
14 Ye, D. Z. , S. W. Luo, B. Z. Zhu, 1957: The wind structure and heat balance in the lower troposphere over Tibetan Plateau and its surroundings, Acta Meteor. Sinica (in Chinese), 28, 108-121.
15 Zhang, Y. W. , D. H. Wang, P. M. Zhai, et al. , 2012: Applicability of AIRS monthly mean atmospheric water vapor profiles over the Tibetan Plateau region, J. Atmos. Oceanic Technol. , 29, 1617-1628.
16 Zhang, Y. W. , D. H. Wang, P. M. Zhai, et al. , 2013: Spatial distributions and seasonal variations of tropospheric water vapor content over the Tibetan Plateau, J Climate, doi: 10. 1175/JCLI-D- 12-00574. 1.