Big Data Potential to Support Agricultural Statistics in Indonesia

Agriculture has an important role in achieving the Sustainable Development Goals (SDGs), especially in the first and second goals, namely eradicating poverty and ending hunger.

The agricultural sector also plays an important role in the structure of the national economy. Thus, the provision of timely and quality agricultural statistics is urgently needed not only for policy makers, but also for researchers and farmers in general.

So far, the Central Statistics Agency has carried out its duties as an agency providing quality agricultural statistical data to meet the needs of data users.

One agricultural statistic produced by BPS is data on harvested area obtained through the Area Sampling Framework survey, or commonly known as KSA.

The KSA survey is the result of a collaboration between BPS and the Agency for the Assessment and Application of Technology (BPPT) in an effort to improve the quality of harvest data.

KSA has been implemented since 2018 for rice commodities and 2019 for corn commodities. The basic principle of KSA is to estimate the area by means of direct observation of the condition of land cover every month at the selected observation points as samples. The conditions observed were in the form of plant growth phases at the observation point. In this way, the ASF method is able to estimate the area according to the growth phase and is able to calculate the potential harvest for the next few months with high-quality data. However, like survey methods in general, KSA requires a lot of resources, both human resources, costs, and time required. On the other hand, advances in information technology make big data potential as a source of information in addition to data from field surveys.

One big data that can be utilized is remote sensing data from satellite imagery. Remote sensing data from satellite imagery can be used to support various interests and present information in the agricultural sector. This is due to several advantages, namely that it can be obtained quickly, some satellite data is available for free, there is historical data, and it can be used to observe areas with difficult access.

In agriculture, satellite data can be used for land type classification, land area estimation, and land use mapping. One of the countries that have used remote sensing technology to support agricultural statistics is China. In 2020, China through the National Bureau of Statistics of China (NBS) has developed a system of estimating agricultural land area using remote sensing and sample surveys to generate the planted area of ​​major crops at the provincial and regional levels (Pan et. al. 2012). Remote sensing technology is used to build a sample frame and update the sample frame. This shows that remote sensing technology has a good opportunity to become part of the agricultural statistics business process.

In Indonesia itself, several researchers have developed a prediction model for agricultural land by integrating KSA data and satellite image data. Triscowati (2019) classified the nine phases of rice growth by utilizing KSA data from Banyuwangi Regency and Landsat-8 data using random forest. Marsuhandi (2020) conducted a land classification in Brebes Regency by utilizing KSA and Landsat-8 data, then applied it to predict the rice harvest area in Brebes Regency. Tamara (2021) developed a machine learning model to predict the growth phase of rice. Muchisha (2021) conducted a mapping of potential land for corn farming by utilizing Landsat-8 satellite data.

In addition to the advantages, of course, satellite data also has some drawbacks in its application.

One of them is a large and complex data structure that requires skilled human resources in data processing.

In addition, some satellites that have more detailed spatial resolution are not available free of charge, so they still require a fee if you want to use the data.

However, with all the advantages or disadvantages in its utilization, satellite data still has a good opportunity to be part of the agricultural statistics business process in Indonesia. (Jambi/BPS Kerinci data)