The harmonization of climate data quality control monitoring with stations network in uganda

ABSTRACT

While climatologists are often not directly involved in the routine operations of observational networks, they are key stakeholders group with respect to their outputs. The climatologists, therefore have a duty to ensure that monitoring and feedback systems are in place to help deliver data outputs according to the required standards.

The EIN2001(Environment Information Network) project was responsible to generate information on different environmental layers such as soil, bio-diversity, agriculture, land use and climate layers etc. The

project helped to retool and increase on the density of the monitoring stations in Uganda. The Uganda meteorological department, being one of the stakeholders, took responsibility to improve on climate information generation. During the project the data quality was harmonized taking into account the stations meta data.

The district officials were sensitized on the importance of the records so as to respect the installations. While it was easy to make innovations of Stevenson screen, it was difficult to get enough proper measuring cylinders to cover all stations. German 200cm² cylinders were provided to the stations and a formula for conversion to the standard cylinder of 12.7cm² was instituted assuming the volume of water captured is constant and height (h) to change due to base diameters of the cylinders.

200h*4/(12.7^2p )=F

F= Rainfall as measured in the 12.7²

h = fall captured in the 200cm²

The inehomogenity in the data sets was tested comparing actual data with the station means and other statistical methods were also used. The sparse stations network and rainfall spatial variability in this region could not allow good checks of comparison with the neighbouring station. A consistence test was performed on stations issued new measuring cylinders. Double mass curve analysis showed low cross correlation between stations hence the need for supplementary checks based on the meta data. Rainfall data for Namulonge, a long term observing station was regressed with recordings from Automatic Weather Station (AWS) installed at the station and R was not small but more work on this regression is being done to get conclusive results

Some of meta data which was not previously recorded any where on the station records was obtained by interviewing some of the long time observers and this helped to explain:

• The anomalies and the suspicious zeros in the data sets.
• The sharp and abrupt discontinuity in the data due to changes in the units of measurements.
• War and social strife in the countryside had serious impact on the datasets.

The brutality of weather and changing climate made the rural people willing partners in monitoring the weather. Increased collaboration between the UMD and the district officials, will improve on data coverage and quality. All in all, the quality and quantity of the data collected in Uganda have greatly declined because of nonstandard weather instruments and poor equipment In this case, the data must be rigorously quality controlled before it can reliably be used. The budget constraints in the LDCs cannot support the purchase of standard instruments which are also not available on the local market.

( Neil Plummer et ela WCDMP-No.5 2 2003),