Perbandingan Akurasi Metode Interpolasi Inverse Distance Weighting dan Ordinary Kriging Sebaran Sifat Kimia Tanah Sawah di Kabupaten Bekasi

Abstract

Lahan sawah merupakan lahan pertanian yang sangat produktif untuk komoditas padi, namun produktivitasnya sangat dipengaruhi oleh sifat kimia tanah seperti pH, C-org, N-Total, P-Tersedia, K-dd, Na-dd, dan KTK. Sampel tanah sawah diambil dari 33 titik lokasi yang tersebar di wilayah Kabupaten Bekasi secara komposit dan dianalisis secara kimia. Berdasarkan keterbatasan titik pengambilan sampel, dilakukan metode interpolasi spasial untuk memperkirakan sebaran sifat kimia tanah di seluruh wilayah Kabupaten Bekasi. Penelitian ini bertujuan untuk membandingkan akurasi metode interpolasi Inverse Distance Weighting (IDW) dan Ordinary Kriging (OK) dalam memetakan sebaran spasial sifat kimia tanah sawah di Kabupaten Bekasi serta menganalisis pengaruh elevasi terhadap distribusi sifat kimia tanah sawah. Analisis dilakukan melalui uji statistik deskriptif, korelasi spearman, serta interpolasi spasial menggunakan IDW power (1, 2, 3, 4, dan 5) dan OK dengan model semivariogram (Spherical, Exponential, dan Gaussian). Evaluasi akurasi dilakukan menggunakan Root Mean Square Error (RMSE). Hasil penelitian menunjukkan bahwa sifat kimia tanah memiliki tingkat variabilitas yang berbeda. Sifat kimia tanah menunjukkan korelasi negatif dengan elevasi, yang berarti semakin tinggi elevasi maka kandungan kimia tanah semakin menurun. Pola sebaran spasial menunjukkan wilayah utara dengan elevasi yang rendah didominasi sawah irigasi memiliki kandungan sifat kimia tanah sawah lebih tinggi dibandingkan wilayah selatan yang didominasi sawah tadah hujan dengan elevasi yang semakin tinggi. Metode IDW dengan power 1 pada sebagian parameter dan power 3 pada KTK merupakan metode interpolasi terbaik dalam penelitian ini karena mampu merepresentasikan variasi spasial sifat kimia tanah dengan tingkat kesalahan yang lebih rendah.

Paddy fields are highly productive agricultural lands for rice cultivation, however their productivity is strongly influenced by soil chemical properties, including pH, organic carbon (C-org), total nitrogen (N-total), available phosphorus (P-available), exchangeable potassium (K-dd), exchangeable sodium (Na-dd), and cation exchange capacity (CEC). Soil samples were collected compositely from 33 locations distributed across Bekasi Regency and analyzed for their chemical properties. Due to the limited number of sampling points, spatial interpolation methods were applied to estimate the distribution of soil chemical properties throughout the study area. This study aimed to compare the accuracy of the Inverse Distance Weighting (IDW) and Ordinary Kriging (OK) interpolation methods in mapping the spatial distribution of soil chemical properties in paddy fields of Bekasi Regency and to analyze the influence of elevation on their distribution. The analyses included descriptive statistical analysis, Spearman correlation analysis, and spatial interpolation using IDW power values (1, 2, 3, 4, and 5) and OK with semivariogram models (Spherical, Exponential, and Gaussian). Interpolation accuracy was evaluated using the Root Mean Square Error (RMSE). The results showed that soil chemical properties exhibited different levels of variability. Most soil chemical properties were negatively correlated with elevation, indicating that their concentrations tended to decrease as elevation increased. The spatial distribution patterns revealed that the northern region, characterized by lower elevations and predominantly irrigated paddy fields, had higher levels of soil chemical properties than the southern region, which is dominated by rainfed paddy fields and higher elevations. IDW with a power value of 1 for most parameters and a power value of 3 for CEC was identified as the best interpolation method in this study because it was able to represent the spatial variability of soil chemical properties with lower prediction errors.