GPS Meteorology in Japan and its Implications to VLBI

Ryuichi Ichikawa (richi(AT)

Kashima Space Research Center
Communications Research Laboratory
893-1 Hirai, Kashima, Ibaraki 314, Japan

At present, GPS meteorology are considered to be a powerful method for an understanding of atmospheric variations (Bevis et al., 1994). Moreover, we expect that GPS meteorology will also contribute an improvement of the precision of space geodetic techniques.

Japanese cooperate study supported by Science and Technology Agency of Japan, which is titled "Establishment of Water Vapor Information Database by GPS Meteorology and its Applications to Numerical Weather Prediction, Earthquake Prediction Research and Hydrology", have been started as feasibility study in 1996 Japanese budget year. The main objective of the study is to improve the precision of both the numerical prediction for mesoscale phenomena and the detection of crustal movement. Achievement of the objective is attributed to the interaction of two systems: the dense permanent network of GPS sites of Geographical Survey Institute (GSI) and the numerical prediction data with high resolution (20 km grid intervals and 36 vertical levels) of Japan Meteorological Agency (JMA). As of April 1996, there are 610 GPS sites of GSI. By the end of 1996 fiscal year, 250 GPS sites will be added to the present network.

Ohtani et al. (1996) preliminarily reported the estimation of precipitable water vapor (PWV) by GPS. They confirmed that PWV can be estimated within 3 - 4 mm rms by GPS comparing with radiosonde data sets as shown in Figure 1. However, there are several study fields which must be carried out to realize the main objective. These are planning in the cooperate study as follows.

Figure 1. {Comparisons of PWV obtained by routine radiosonde data sets of JMA (circles and squares) and that estimated from GPS analyses (solid lines) at Sapporo, Akita, Wajima, and Yonago (after Ohtani et al., 1996)

We are now planning to evaluate the tropospheric path delay correction based on numerical prediction data at KSP (The Key Stone Project) sites of Communications Research Laboratory (CRL). KSP network is carried out to monitor the crustal movements around Tokyo metropolitan area, Japan. The network will be equipped with three space geodetic facilities, i.e. VLBI, SLR, and GPS, routinely operated by the end of 1996. Precise correction of tropospheric path delay is indispensable to detect seismic precursors and coseismic movements by using KSP network.

VLBI geodesy can avoid uncertainties of satellite orbit determinations and can independently estimate the tropospheric path delay at each site by long distance measurements. Thus, it is expected that the residual delay estimated by VLBI will be more precise using the correction based on numerical prediction data. The estimated delay is available in the analysis of GPS data as a priori delay. As a result, we can reduce an unknown parameter in the GPS analysis. In addition, realtime VLBI analysis is now under development. Consequently, more precise monitoring of crustal movements by VLBI and GPS will be carried out with sub-daily temporal resolution.

Figure 2. Time series of zenith tropospheric delay estimated by VLBI and GPS analyses at Koganei - Kashima baseline (about 109 km) during 8 - 14 July, 1996. Since a priori delay is not used in this analyses for both VLBI and GPS, the absolute values are not confident. However, similar characteristics are shown for both plots by VLBI and GPS; increasing in Exp.2 (July 9) and Exp. 3 (July 10), and decreasing in Exp. 4 (July 11).

We are now executing a feasibility study for estimating the delay by KSP network and are preparing for observations comparing with water vapor radiometer (WVR). Figure 2 shows time series of the zenith delay estimated by VLBI and GPS analyses at Koganei and Kashima during 8 - 14 July, 1996 (JST). Both plots by VLBI and GPS increase during 9 - 11 July (JST) and extremely decrease during 11 - 12 July (JST) in a similar manner. According to the infrared images by a geostationary meteorological satellite of JMA (GMS5), a typhoon moved northward in the Western Pacific during the concerned period (see Figure 3). It seems that the variations of the zenith delay are attributed to the water vapor variation caused by the typhoon movement though it is a roughly qualitative comparison. For more detailed interpretation another investigations are required; comparisons with numerical prediction data and radiosonde data sets, recalculations of parameter estimations in VLBI and GPS analyses, and so on. The results of these investigations will be reported in another paper.

Figure 3. Infrared images by GMS5 of JMA. Arrows as shown in these figures indicate typhoon. It is inferred that the variation of water vapor related to the typhoon causes the temporal variations of zenith delay as shown in Figure 2.


Bevis, M., S. Businger, S. Chiswell, T. A. Herring, R. A. Anthes, C. Rocken, and R. H. Ware (1994): GPS meteorology: Mapping zenith wet delays onto precipitable water, J. Appl. Meteor., 33, 379-386.

Ohtani, R., H Tsuji, N. Mannoji, J. Segawa, and I. Naito (1996): Precipitable Water Vapor Observed by Geographical Survey Institute's GPS network, Tenki, in submitting (in Japanese).

Updated on October 28, 1996. Return to CONTENTS