Analisis dan Pemodelan Emisi Gas Rumah Kaca Padi Sawah Irigasi dengan Sistem Alternate Wetting-Drying (AWD)

Abstract

Budidaya padi sawah merupakan salah satu sumber utama emisi gas rumah kaca (GRK), terutama metana (CH4) dan dinitrogen oksida (N2O). Pengelolaan air melalui sistem Alternate Wetting and Drying (AWD) berpotensi meningkatkan produktivitas sekaligus menekan emisi GRK. Penelitian ini bertujuan menganalisis pengaruh berbagai perlakuan irigasi terhadap produktivitas padi, mengembangkan model pertumbuhan tanaman, mengevaluasi emisi GRK dan Global Warming Potential (GWP), serta mengembangkan model Artificial Neural Network (ANN) untuk estimasi emisi GRK. Penelitian menggunakan tiga perlakuan irigasi, yaitu Continuous Flooding (FL), Wetting (WT), dan Drying (DR). Model pertumbuhan dikembangkan menggunakan model logistik Verhulst, sedangkan model emisi GRK menggunakan ANN dengan lima parameter lingkungan tanah sebagai variabel input. Hasil penelitian menunjukkan bahwa perlakuan DR menghasilkan produktivitas tertinggi sebesar 10,75 t ha-1, sedangkan FL menghasilkan produktivitas terendah sebesar 7,79 t ha-1. Emisi CH4 dan GWP terendah juga diperoleh pada DR, masing-masing sebesar 82,50 kg ha-1 dan 2.539 kg CO2-eq ha-1, atau menurun 76,2% dan 74,2% dibandingkan FL. Dari sisi penggunaan air, perlakuan WT menghemat air irigasi sebesar 4,40% dan DR sebesar 12,04% dibandingkan FL. Hasil ini menunjukkan bahwa penerapan DR mampu meningkatkan produktivitas padi sekaligus efisiensi penggunaan air tanpa menghambat pertumbuhan tanaman. Model logistik mampu menggambarkan pertumbuhan tanaman dengan baik (R2 = 0,894 - 0,996), sedangkan model ANN menunjukkan akurasi yang tinggi dalam memprediksi emisi CH4 (R2 = 0,92) dan N2O (R2 = 0,82). Hasil penelitian menunjukkan bahwa sistem AWD melalui perlakuan DR efektif meningkatkan produktivitas padi sekaligus menurunkan emisi GRK, serta bahwa model ANN berpotensi digunakan sebagai alat prediksi emisi GRK pada budidaya padi sawah berkelanjutan.

Rice cultivation is one of the major sources of greenhouse gas (GHG) emissions, particularly methane (CH4) and nitrous oxide (N2O). Water management through the Alternate Wetting and Drying (AWD) system has the potential to improve rice productivity while reducing GHG emissions. This study aimed to analyze the effects of different irrigation treatments on rice productivity, develop crop growth models, evaluate GHG emissions and Global Warming Potential (GWP), and develop an Artificial Neural Network (ANN) model for estimating GHG emissions. The experiment employed three irrigation treatments, namely Continuous Flooding (FL), Wetting (WT), and Drying (DR). Crop growth was modeled using the Verhulst logistic growth model, while GHG emissions were estimated using an ANN model with five soil environmental parameters as input variables. The results showed that the DR treatment produced the highest grain yield of 10.75 t ha-1, whereas FL resulted in the lowest yield of 7,79 t ha-1. The lowest CH4 emissions and GWP were also observed under DR, reaching 82.50 kg ha-1 and 2,539 kg CO2-eq ha-1, respectively, representing reductions of 76.2% and 74.2% compared with FL. In terms of water use, the WT treatment saved 4.40% of irrigation water and 12.04% of DR compared to FL. These results indicate that DR can increase rice productivity and water efficiency without inhibiting plant growth. The logistic model successfully described crop growth with coefficients of determination (R2) ranging from 0.894 to 0.996. Meanwhile, the ANN model demonstrated high predictive accuracy for CH4 (R2 = 0.92) and N2O (R2 = 0.82) emissions. The findings indicate that the AWD system through the DR treatment is an effective water management strategy for increasing rice productivity while reducing greenhouse gas emissions. Furthermore, the developed ANN model shows strong potential as a tool for predicting GHG emissions in sustainable irrigated rice cultivation systems.