Analisis Performa Pemodelan Iklim Global Dalam Prediksi Awal Musim di Sulawesi
Performance Analysis of Global Climate Models for Predicting the Onset of Seasons in Sulawesi
Keywords:
SEAS5; CFSv2; seasonal onset; evaluation; HSSAbstract
Keberagaman tipe Zona Musim (ZOM) di sulawesi, sebagai wilayah dengan tipe ZOM terbanyak, memberikan tantangan dalam memprediksi awal musim. Penelitian ini mengevaluasi dan membandingkan performa model ECMWF SEAS5 dan CFSv2 dalam memprediksi awal musim hujan (AMH) dan awal musim kemarau (AMK) di berbagai tipe ZOM di Sulawesi pada periode tahun 1991-2020. Data yang digunakan berupa data prediksi curah hujan SEAS5 dan CFSv2, yang telah dikoreksi bias, serta data observasi CH Blending. Analisis dilakukan dengan mengelompokkan ZOM ke dalam 10 tipe klimatologis, kemudian menentukan Lead Time (LT) berdasarkan normal awal musim tiap kelompok. Evaluasi performa model dihitung menggunakan metode Multi-Category Contingency Table (MCCT) yang dilengkapi dengan perhitungan akurasi dasarian (ACC1) dan dasarian±1 (ACC2) serta perhitungan skill menggunakan Heidke Skill Score (HSS).
Hasil penelitian ini menunjukkan bahwa secara umum model SEAS5 lebih baik dalam memprediksi awal musim dibandingkan model CFSv2. Nilai akurasi tertinggi dicapai oleh SEAS5-corr, dengan akurasi AMH tertinggi pada ZOM Monsunal-2A inisial prediksi September (90-100%), serta akurasi AMK tertinggi pada Monsunal-2B dan Ekuatorial-4A pada inisial September (80-90%). Sebaran nilai HSS menunjukkan pola yang sesuai dengan akurasi, di mana ZOM Monsunal dan Ekuatorial-2 menghasilkan skill terbaik yang dominasi nilai positif. Sebaran HSS tinggi, lebih besar dari 0.50, dijumpai pada ZOM Monsunal-2A, monsunal-2B dan Ekuatorial-4A. Secara umum, kedua model menunjukkan performa yang lebih baik pada pendekatan ACC2 dan LT pendek. Model SEAS5-corr menunjukkan performa yang lebih unggul dibandingkan CFSv2-corr, kecuali pada AMK Lokal-4 dan AMH Ekuatorial-4B.
Kata kunci: SEAS5; CFSv2; awal musim; evaluasi; HSS
The diversity of Seasonal Zone (ZOM) types in Sulawesi, as the region with the highest number of ZOM in Indonesia, poses a challenge in predicting seasonal onset. This study evaluates and compares ECMWF SEAS5 and CFSv2 model performance in predicting the rainy season onset and dry season onset across various ZOM types in Sulawesi for the 1991-2020 period. The data used includes bias-corrected SEAS5 and CFSv2 precipitation prediction, and blended observed precipitation. The analysis was conducted by grouping the ZOMs into 10 climatological types, followed by determining the Lead Time (LT) based on the normal seasonal onset of each group. Model performance evaluation was calculated using Multi-Category Contingency Table (MCCT) method, complemented by exact ten-day accuracy (ACC1) and ten-day±1 (ACC2) approach, and Heidke Skill Score (HSS).
The result show that both models are generally capable to represent the climatological patterns of ZOM in Sulawesi, although SEAS5-corr predictions are more representative. The highest accuracy was achieved by SEAS5-corr, with the highest AMH accuracy found in Monsoonal-2A for initial month in September (90-100%), and the highest AMK accuracy in Monsoonal-2B (July and September initialization) and Equatorial-4A (September initialization) at 80-90%. The distribution of HSS values aligns with the accuracy patterns, where Monsoonal and Equatorial-2 shows the best skills, dominated by high values. Highest HSS values were observed exceeding 0-50 in Monsoonal-2A/2B and Equatorial-4A. In general, both models perform better using the ACC2 approach and short LT. Overall, the SEAS5-corr model consistently outperformed CFSv2-corr, except for AMK in Local-4 and AMH in Equatorial-4B.
Keywords: SEAS5; CFSv2; seasonal onset; evaluation; HSS
References
Aldrian, E., & Susanto, R. D. (2003). Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. International Journal of Climatology, 23(12), 1435–1452.
Badan Meteorologi, Klimatologi, dan Geofisika. (2022). Pemutakhiran zona musim Indonesia periode 1991–2020. Pusat Informasi Perubahan Iklim, Kedeputian Bidang Klimatologi.
Heidke, P. (1926). Berechnung der erfolges und der güte der windstärkevorhersagen im sturmwarnungdienst. Geografiska Annaler, 8, 301–349.
Johnson, J. A. (2012). Assessing the impact of climate change in Borneo. World Wildlife Fund.
Johnson, S. J., Stockdale, T. N., Ferranti, L., Balmaseda, M. A., Molteni, F., Magnusson, L., & Mogensen, K. (2019). SEAS5: The new ECMWF seasonal forecast system. Geoscientific Model Development, 12(3), 1087–1117.
Jolliffe, I. T., & Stephenson, D. B. (2003). Forecast verification: A practitioner's guide in atmospheric science. Wiley.
Kurnia, W., Muharsyah, R., & Widiyanto, S. (2020). Performa koreksi bias prakiraan curah hujan model European Centre Medium Weather Forecast (ECMWF) di Sulawesi. Buletin GAW Bariri, 1(2), 77–86.
Marjuki, M., Koesmaryono, Y., Santikayasa, I. P., & Sopaheluwakan, A. (2025). Assessing the influence of climate services and climate change adaptation strategies on smallholder agriculture: A systematic literature review. Agromet, 39(2), 75–85. https://doi.org/10.29244/j.agromet.39.2.75-85
Miokta, T. G. (2023). Evaluasi model prediksi musim global menggunakan model ECMWF SEAS5 untuk penentuan awal musim berbasis ZOM9120 di Pulau Jawa [Skripsi, Sekolah Tinggi Meteorologi Klimatologi dan Geofisika].
Muharsyah, R. (2017). Kajian verifikasi produk prakiraan curah hujan bulanan (2003–2012). Jurnal Meteorologi dan Geofisika, 18(2), 51–59.
Muharsyah, R., Ripaldi, A., Maharani, T., Fitrianti, N., Hanif D., R., Denata, M., & Jauhari. (2020). Perbandingan model kopel ECMWF System 4 dan CFSv2 untuk prediksi musim di Indonesia. Megasains, 11(1), 1–11.
Murphy, A. H. (1993). What is a good forecast? An essay on the nature of goodness in weather forecasting. Weather and Forecasting, 8(2), 281–293.
National Academies of Sciences, Engineering, and Medicine. (2010). Assessment of intraseasonal to interannual climate prediction and predictability. The National Academies Press.
Naylor, R. L., Battisti, D. S., Vimont, D. J., Falcon, W. P., & Burke, M. B. (2007). Assessing risks of climate variability and climate change for Indonesian rice agriculture. Proceedings of the National Academy of Sciences, 104(19), 7752–7757. https://doi.org/10.1073/pnas.0701825104
Rohmah, I. A. S. (2025). Evaluasi performa model ECMWF SEAS5 dalam prediksi awal musim di Kalimantan berbasis ZOM 1991–2020 [Skripsi, Sekolah Tinggi Meteorologi Klimatologi dan Geofisika].
Safi’i, A., & Sopaheluwakan, A. (2020). The skill of CFSv2 in predicting rainfall anomalies over Indonesia during different ENSO phases. AIP Conference Proceedings, 2223(1), 020005.
Saha, S., Moorthi, S., Wu, X., Ji, J., Nadiga, S., Tripp, P., Behringer, D., Hou, Y., Chuang, H., Iredell, M., Ek, M., Meng, J., Yang, R., Mendez, M. P., van den Dool, H., Zhang, Q., Wang, W., Chen, M., & Becker, E. (2014). The NCEP Climate Forecast System Version 2. Journal of Climate, 27(6), 2185–2208.
Sanjaya, W. (2023). Evaluasi curah hujan model ECMWF SEAS5 di Indonesia [Skripsi, Sekolah Tinggi Meteorologi Klimatologi dan Geofisika].
Supari, S., et al. (2022). Pemutakhiran zona musim (ZOM) berdasarkan normal curah hujan 1991–2020. Badan Meteorologi, Klimatologi, dan Geofisika.
Wang, Q. J., Shao, Y., Song, Y., Schepen, A., Robertson, D. E., Ryu, D., & Pappenberger, F. (2019). An evaluation of ECMWF SEAS5 seasonal climate forecasts for Australia using a new forecast calibration algorithm. Environmental Modelling & Software, 122.
