<![CDATA[Stasiun pengamatan cuaca Badan Meteorologi, Klimatologi, dan Geofisika (BMKG) menggunakan data pengukuran suhu udara pada pukul 07.00 (T07), 13.00 (T13), dan 18.00 (T18) waktu setempat setiap hari untuk keperluan data sinoptik meteorologi/klimatologinya. Data tersebut ditetapkan sebagai Kelompok Data Utama (KDU) guna menentukan suhu udara rata-rata harian (operasional), berdasarkan formula Top = ½ T07 + ¼ T13 + ¼ T18, di mana bilangan ½, ¼, dan ¼ adalah Kumpulan Nilai Koefisien (KNK) dengan urutan dan nilai masing-masing komponennya menyatakan bobot yang bersesuaian dengan urutan dari komponen data pada KDU terkait. Perumusan ulang formula di atas menggunakan metoda Least Square ternyata memunculkan lebih banyak pilihan KNK yang mungkin. Sementara itu, dari data suhu udara tiap jam yang terukur pada Automatic Weather Station (AWS) di Stasiun Klimatologi BMKG Kota Palembang selama tahun 2007 dan dari Desember 2010 sampai dengan Nopember 2011 formula Least square mengusulkan KNK = {0.46, 0.23, 0.31} yang, bila dibandingkan dengan nilai rata-rata sebenarnya, menghasilkan nilai suhu udara rata-rata harian lebih baik daripada hasil perhitungan dengan formula di atas, dengan koefisien korelasi = 0.87±0.05 terhadap 0.85±0.06, simpangan rata-rata = (0.01±0.09)°C terhadap (0.03±0.09)°C, simpangan mutlak rata-rata = (0.35±0.05)°C terhadap (0.37±0.06)°C, dan simpangan akar kuadrat rata-rata = (0.44±0.07)°C terhadap (0.47±0.08)°C.Agency for Meteorology, Climatology, and Geophysics (BMKG) weather stations use three times daily air temperature measurements at 07.00 (T07), 13.00 (T13), and 18.00 (T18) local times for their synoptic meteorological or climatological data. These measured data are assigned as a Primary Data Set (KDU) and used to determine (operational) daily mean air temperature (Top,) based on a formula (mentioned as BMKG formula) Top = ½ T07 + ¼ T13 + ¼ T18, where ½, ¼, and ¼ are a Collection of Coefficient Values (KNK). The orders and values of each component of KNK state weights which are suitable to the sequence of data components from related KDU. Reformulation of the formula above using the least square method gave more options for possible KNK. Applying the least square method to hourly air temperature data measured by Automatic Weather Station (AWS) in Palembang Climatological Station during 2007 and from December 2010 to November 2011 produced a KNK of {0.46, 0.23, 0.31} which gave values of daily mean air temperature closer to the true mean than the Top obtained by formula above (BMKG formula). Comparisons of correlation coefficients, mean errors, absolute mean errors and root mean square errors between the results of a proposed formula against a BMKG formula were: 0.87±0.05 vs 0.85±0.06; (0.01±0.09)°C vs (0.03±0.09)°C; (0.35±0.05)°C vs (0.37±0.06)°C; and (0.44±0.07)°C vs (0.47±0.08)°C, respectively. ]]>
Copyrights © 2015
