Handheld GPS equipment for agricultural statistics surveys
Experiments on area-measurements done during fieldwork for the Uganda Pilot Census of Agriculture, 2003
Uganda Bureau of Statistics and Statistics Norway staff jointly conducted fieldwork in Uganda 2003 experimenting with simple handheld Geographical Positioning System (GPS) tools for determination of agricultural holding areas and geo-referencing of holdings. NORAD funded this exercise under the "Support to Strengthen Agricultural Statistics" Program 2002-2005. The objective of the exercise was to test out the feasibility of using a hand held Global Positioning System (GPS) tool as an alternative instrument for area measurement during agricultural fieldwork in Uganda. The experiment was conducted as a part of the Pilot Agricultural Census (PCA). More than 900 holdings distributed in 10 districts of Uganda were visited and parcel and crop plot areas where measured using several methods. Therefore a substantial number of observations are available for comparative analyses. From the PCA there are a total of 430 observations where areas of parcels were measured both with GPS and by traversing (tape & compass). Tape and compass measurements is regarded as the most accurate observation of the ground truth taken during the experiment. However, a paired T-test of this set of observations reveals that there is minor difference between the results of GPS use compared to Traversing concerning parcels measured during the PCA fieldwork 2003. The same conclusion was also drawn when measurement of the smaller crop plot areas were compared. For both parcel and crop plot areas, the farmers and enumerator's eye-estimates were found not to be reliable. A subset of 191 observations about time use on holdings where both GPS measurements and traversing with tape and compass was conducted. It can be observed that the average time use per holding for traversing was as much as 3 hours and 23 minutes or 3.5 times as much as when GPS was used. The price of high quality tape and compass equipment was, at the time of the experiment, approximately 25 USD and 100 USD (compass including jacket) respectively. In addition a fairly expensive programmable calculator is necessary to calculate areas captured by traversing. The total price is therefore not so different from the price of the simple handheld GPS tool used. However, battery costs are high for GPS use, while almost neglectable for traditional traversing.
Three recommendations from this study are presented as follows: Recommendations 1 - findings from the PCA indicate that there is potential to use relatively cheap Global Positioning System (GPS) equipment for measuring of area and for geo-referencing of holdings in the context of agricultural statistics. Recommendation 2 - GPS measurement without additional equipment for adjustments of signals and/or improved antenna proves to give some variation in the measured area. Therefore, repeated measurement and calculation of an average of the same area as well as further improvement of tools set-up and methods, is recommended. Recommendation 3 - Training of fieldworkers in the technical use of the GPS, including how to set it up correctly, is crucial.
Acknowledgement : We are grateful to NORAD for its funding of the SSASP project and thereby enabling for improvement of the agricultural statistics in Uganda by further development of methods and introduction of new technology. Especially thanks to all the Ugandan farmers, fieldworkers and entry staff contributing to this exercise.