Aquacrop is free-licensed Food and Agricultural Organization’s crop modelling that requires minimum inputs of climate variables namely rainfall, maximum temperature, minimum temperature variables and geographic information of the area to be simulated (longitude, latitude, altitude). This study aims to measure the difference in irrigated and rainfed rice productivity from the Aquacrop crop modelling simulation to the influence of climate pattern variations in Java Island, Indonesia. The k-means clustering method applied to the rainfall, maximum, and minimum temperature variables from the bias-corrected MERRA2 data resulted in two climate regions. The principal component analysis result showed that the maximum and minimum temperature variables are the variables that most contribute to the determination of the clustering area using the k-means method compared to the rainfall variable. This study has calculated the probability of the irrigated and rainfed rice productivity resulting from the Aquacrop simulation in those climate regions during La Nina [El Nino] years that will be higher [smaller] than the mean value of rice productivity during neutral years. However, the validation between the actual irrigated and rainfed rice productivity with the Aquacrop simulation results from 2001-2014 showed low correlation values that vary between negative and positive values in all climate regions. Meanwhile, the validation on the El Nino composite years generally showed positive correlation values. In addition, the neutral and La Nina composite years resulted in varying correlation values between negative and positive correlation.

Penelitian ini bertujuan untuk mengetahui perbedaan secara statistik hasil pengukuran suhu udara menggunakan peralatan otomatis (Automatic Weather Station/AWS) dengan hasil pengukuran suhu udara secara manual atau sinoptik. Data yang diuji adalah data per-tiga-jam-an yang berasal dari data AWS dan data sinoptik dari 12 (dua belas) stasiun milik Badan Meteorologi Klimatologi dan Geofisika (BMKG) yaitu Bengkulu,

<p>Large-scale forest fires often occur in Indonesia and affect air quality and human health. The effect of forest fire on air quality quantified by rising PM<sub>10</sub> concentration on Indonesia Meteorological, Climatological and Geophysical Agency (BMKG) observation network. A few PM<sub>10</sub> observation networks and uneven distribution in Indonesia make it difficult to present spatial ground-level PM<sub>10</sub>.  The aim of this study was to estimate ground-level PM<sub>10</sub> in Indonesia and present the spatial distribution of ground-level PM<sub>10</sub> using machine learning. Support Vector Regression (SVR) techniques were used to estimate the PM<sub>10</sub> content from heterogeneous data sources, including ground measurements provided by BMKG, numerical model data, and hotspot retrieved from NASA/LANCE – FIRMS for satellite imagery. RMSE and MSE were used to evaluate the estimation result. We also present the modeling framework on the forecast of the CAMS Copernicus model in Indonesia. The performance of various input parameter configurations of SVR for estimating the ground-level PM<sub>10</sub> as indicated by low prediction errors.</p>

This study aims to compare the relationship between climate variables and rice productivity under different irrigation systems (irrigated and rainfed) in the clustering area on Java Island, Indonesia. This study used the clustering areas resulting from the previous study. The climate variables used are bias-corrected MERRA2 data from the period 1987–2017, cropped for Java Island. The rice productivity and reference evapotranspiration data used in this study are the results of the simulation of Aquacrop modeling. The result from the cluster method used tends to divide Java Island into 2 clusters with different altitudes (lowland and highland) areas. The results show that the correlation values between the precipitation variable and rice productivity from Aquacrop simulation (both irrigated or rainfed) in cluster 1 (dominated lowlands) are higher than in cluster 2 (dominated highlands), contrary to that the correlation values between the reference evapotranspiration variable with rice productivity from Aquacrop simulation (both irrigated or rainfed) are higher in cluster 2 (dominated highlands) areas, compared to cluster 1 areas (dominated lowlands). R-square values from response surface methodology (RSM) on the rainfed system in both clusters are higher than those on the irrigated system. This indicates that rainfed agriculture is highly dependent on climate variables, especially precipitation and reference evapotranspiration variables compared with the regular irrigated agricultural system. The RSM result also shows that climate variables significantly contribute to the variation of rice productivity generated by Aquacrop modeling in irrigated and rainfed systems and in all clusters.