DIGITALNA ARHIVA ŠUMARSKOG LISTA
prilagođeno pretraživanje po punom tekstu



ŠUMARSKI LIST 7-8/2016 str. 47     <-- 47 -->        PDF

have been estimated by GAPS service solutions. When comparing relative solutions for TP01, it is found that AUSPOS service gives more accurate solutions than OPUS service. Moreover, the standard deviations obtained from OPUS service are bigger than AUSPOS service solutions.
The coordinate differences and standard deviations of TP02 computed by Eq. 3 and Eq. 4 are represented graphically as shown in Figure 7-8-9 for Doy: 198, 199 and 200, respectively. TP02 is located under the forest are where the satellite visibility is problematic. In this condition, it is seen that web based online services using PPP solution give better accuracy than the services using relative technique. Having small differences on the results, it can be also concluded that GAPS service gives more reliable solutions for TP02 with respect to standard deviations as well. Also with respect to the standard deviations, standard deviations estimated from relative technique are bigger than estimated from services using the PPP solution. This is a significant factor in terms of accuracy of the services. The reason of estimating more reliable solutions from the services using PPP technique is that they are able to model most of the error sources that can be occurred during the observation by considering more parameters. Modeling the errors as listed on Table 3 is performed automatically with the related services. However, some of the services present alternatives of error models that give preferences to the users. Advanced mode of GAPS service can be given as an example of this situation.
While comparing the results for TP01 and TP02, which have different locational characteristic, the effect of satellite visibility on positional accuracy stand out. All solutions provided by services for TP01 is more reliable than TP02 solutions and the positional accuracies are better as well.
At the above figures, differences of each coordinates components have been represented separately. Moreover, in this study, to investigate the accuracy and reliability of these services, 3D positioning differences are computed from Eq. 5. Standard deviations of the positional error are calculated by Gaussian error propagation law.
Naesset and Gjevestad (2008) have collected 2 hours static GPS data in the forest environment and post-processed the data with PPP technique. They have used IGS satellite orbit and clock products in the evaluation process. They have post-processed the data with TerraPos software and from the PPP solution for 2 hours observations; they have provided the mean positional accuracy ranging from 0.270 m to 0.880 m (Naesset and Gjevestad, 2008). While the observation duration is increased to 3 hours, it can be concluded that the results with PPP solution that have been obtained by 2 hours observations, can be more reliable.