Türkiye koşullarında yapılan GPS gözlemlerinde ortam etkilerinin araştırılması

Abstract

Son yıllarda bilimsel amaçlı olarak kullanılan uydu tekniklerinde çok önemli gelişmeler olmuştur. Uydu sistemlerindeki bir çok hata kaynağı ise bu gelişmeler sonucu hemen hemen giderilmiştir, örneğin, uydu saati hataları başlangıçta, uydu gözlemlerinde, önemli bir hata kaynağını oluşturmaktayken bugün özellikle GPS uydularında atomik (Hydrogen Masers) saatlerin kullanılması ile bu hata kaynağı gözardı edilebilecek kadar azaltılmıştır. GPS'in Jeodezik ve Jeodinamik amaçlı uygulamalarında temel olarak uydu- alıcı arasındaki uzaklığın duyarlı olarak belirlenmesi söz konusu olduğundan bu uzaklığın belirlenmesinde etkili olan tüm hata kaynaklarının doğru olarak hesaplanması ya da modellendirilmesi gerekmektedir. Günümüzde GPS, yukarıda belirtilen amaçlar için yoğun bir şekilde kullanılmaktadır. Bilindiği gibi yüksek duyarlık isteyen çalışmalarda göreli konumlama yöntemi ve faz ölçüleri tercih edilmektedir. Uydu tekniklerindeki son gelişmeler sonucu artık bu ölçülerdeki en önemli hata kaynağını ortam etkileri (atmosferik etkiler ve multipath) oluşturmaktadır. Dolayısıyla, GPS ile elde edilen ölçülere Jeodezik ve Jeodinamik amaçlarla kullanılmadan önce mutlaka atmosferik düzeltmeler getirilmelidir. Bu çalışmada ortam etkileri, iyonosferik etki, troposferik etki ve sinyal yansıma (multipath) etkisi olarak üç bölümde incelenmektedir. Yukarıda belirtilen her bir etki koordinat bileşenlerini (enlem, boylam, elipsoit yüksekliği) ve baz uzunluklarını farklı şekilde etkilediğinden her bölüm için ayrı sayısal uygulama yapılmış ve GPS ölçülerinin değerlendirilmesiyle ilgili ölçme ve hesaplama önerileri belirlenmiştir.

In order to obtain high precision measurements via space techniques, it is essential to know the influence of the propagation errors ( atmospheric effects and multipath ) on these measurements. Today, GPS is being used intensively for geodetic and geodynamics purposes. For scientific studies which require high precision the relative static positioning technique using phase observations are preferred. However, in phase observations one of the most important error sources is atmospheric delay. Consequently, atmospheric corrections should be applied to the measurements obtained via GPS before using them for geodetic, geodynamics and astronomical purposes. As it is well-known, GPS (microwave) signals arrive at the receiver passing through the Earth's atmosphere, which not only bends the ray, but also slows it (figure-1). In processing, it is normally the straight line G plus the path length outside the atmosphere which is required. This can be expressed as, G = ]dG (1) XL Figure 1. Atmospheric Effects But in reality this is not the situation. Since the electromagnetic waves follow the minimum electrical path length (Fermat's principle) between two points normally the measured length ; S, can be written as S=|n.dS (2) Here, n is the refractive index of the atmosphere. Since it is important to determine the excess path length in GPS processing, this excess path length ( AL) can be given as, W WW AL = S -G=J(n(s)-1).dS+JdS-JdG (3) There are two reasons for AL: (a) Since the refractive index is greater then unity, this slows the signals coming from the vacuum into the atmosphere, Ml (b) Since the coming signal is bent, they cannot follow a straight line. So, the above equation shows that the error in GPS observation is due to two components. These are, (a) The excess path length due to the propagation delay, J(n(s)-1).dS (b) The excess path length due to bending, b c JdS-JdG a a The second of these is usually not significant except at low observation angles (Table-1). Table-1. Cut-off elevation-bending values relationships If we express the propagation delay in terms of the refractivity, N, of the atmosphere using the relationships XM N=(n-1)x108 (4) We can write the excess path length due to propagation delay as AL = 10"6jN(s)dS (5) When it is taken into consideration that the GPS signals arrive at the receiver passing through the atmosphere, there are two atmosphere layers which affect the GPS observations. They are Ionosphere and Troposphere. Ionosphere is upper of the two layers, extending from about 50 km up to approximately 500 km. Troposphere is the region below this, extending from ground level up to the base of the ionosphere. These two layers, therefore, affect the GPS observations in different ways. Consequently, they are dealt with separately. Since their effects are different on GPS observations, they are taken into consideration using different modelling and techniques in GPS processing. On the other hand, multipath is completely a different error source on GPS measurements. As it is known, GPS receiver antennas are omnidirectional. So, depending on the terrain where the antenna is set up and the chosen cut-off elevation angle undesired satellite signals can be recorded in the receiver. As a definition, multipath is the phenomenon whereby a satellite emitted signal arrives at the receiver via two or more routes. This situation is mainly caused by reflecting surfaces near the receiver. Besides, multipath can occur at both end of the system. These are called antenna and satellite multipath, (Figure-2 and 3). XIV Reflected Path (Multipath) D!rect Path ",,. XAXAXAXAXAXA Receiver Antenna a A A. Figure-2. Antenna Multipath Reflected Path (Multipath) Direct Path Antenna r\ 'X Towards Earth Centre Figure-3. Satellite Multipath Multipath effects cause serious problems for a GPS receiver in an environment of reflecting surfaces. And, there is no general model of the multipath effect due to the arbitrarily different geometric situations. However, the influence of the multipath can be estimated by using a combination of L1 and L2 code and carrier phase measurements. From the geometry it is clear that signals received from low satellite elevations are more susceptable to multipath than signals received from high elevations. And code ranges are more strongly affected by multipath than carrier phases. XV There can be counted many measures against multipath. The most effective measure is to avoid it. That is, it is the best measure to avoid reflecting surfaces as far as possible. Except for the above mentioned measures against multipath there are other methods. For example, using an antenna which takes advantage of signal polarization, or using digital filtering, wide band antennas, or radio frequency absorbent antenna ground planes reduce the interference of satellite signals. For the investigation of ionosphere errors induced on GPS baselines, two long and two short baselines are selected. For the long baselines, GPS observations performed at 3 SLR sites ( Yığılca, Yozgat and Melengiçlik) in 1991 are processed. For the short baselines, GPS observations performed at Ankara GPS Test Network (AGTN) in 1990 are processed. Processings were done using Bernese Software version 3.3 at General Command of Mapping (GCM). Processing results were compared with the SLR values of the long baselines and the EDM distances of the short baselines. As a result it is seen that it is almost better to use L1&L2 strategy with ambiguity fixed solutions for the short baselines (< 20 km ). And for the long baselines ( > 50 km ) it is better to use L3 linear combination strategy with/without ambiguity fixed solutions. For the investigation of tropospheric errors especially on GPS heights several processing strategies were applied. As a result, it became clear that if highly precision and well calibrated meteorological instruments are used for recording surface meteorological data (SM) and by using these measured SM values and estimating residual tropospheric parameters at N points give comparable results. On the other sites, an alternative method is to use a standard atmospheric model and solve for residual tropospheric parameters. In this study it is concluded that both methods are compatible. Hence, in order to obtain tropospheric bias free height values, processing SM data obtained using high quality and expensive meteorological instruments do not lead to more accurate GPS height results than using a standard atmosphere and estimating residual tropospheric parameters. So, it is more economical and easier way to use a standard atmosphere model and estimate residual tropospheric parameters in processing of GPS observations. In order to show the effects of tropospheric errors on GPS observations, real GPS data xvi were used in processings which were collected during the Marmara GPS Campaign performed in 1994 in Turkey with the collaboration of General Command of Mapping (GCM), Technical University of istanbul (İTU) and Technical University of Zurich (ETHZ). This study consists of 6 sections. In the first section introduction is presented. General descriptions and information relevant to atmosphere are given in section 2; and Ionosphere and troposhere are dealth with separately in sections 3 and 4, respectively. Section 5 deals with multipath and in section 6 results are given.