#TITLE_ALTERNATIVE#

• Main Author: SHOFIYATI (NIM. 35107301); Tim Pembimbing: Prof. Ir. Ketut Wikantika, M.Eng., Ph.D.; Prof. D, RIZATUS
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/19341
• Institution: Institut Teknologi Bandung

To strengthen food security government launched a self-sufficiency of five commodities in the period 2010 to 2014, where rice is one of them. Data and accurate information is a very important element for monitoring in agriculture. <br /> <br /> Spatial information about cropping patterns is important for management of agricultural land. Further analysis of data and information is useful to consider a better management planning and development of agricultural land. To establish food security the government announced a target of five commodities self-sufficiency in period 2010 – 2014, where rice is one of them. <br /> <br /> Accurate data and information is a very important element for agriculture monitoring. Spatial information about cropping patterns is important for the management of agricultural land. Further analysis of data and information is useful for considering a better management planning and development of agricultural land. <br /> <br /> The existing rice estimation production which uses statistical estimation methods of eye estimate often raises the pros and cons, because the results of <br /> <br /> calculations tend to be overestimates. To this the Ministry of Agriculture plans to do some new methods to support the existing method. These include acreage <br /> <br /> estimate spatially using satellite data and remote sensing technology. <br /> <br /> Currently, the method of remote sensing based on qualitative analysis (logical inference analysis) to analyze the paddy growth period still has limitation of accuracy, objectivity, and consistency. Similarity of food crops appearance on image data is difficult to identify types of crops and cropping patterns. It is <br /> <br /> necessary to study a new model or method that can distinguish paddy and other crops. <br /> <br /> This research aims to create a new algorithm Composite Orthogonal Transformation (COT) multi spectral and SAR data to develop a model of the cropping pattern and map a cropping pattern distribution in paddy fields in the <br /> <br /> northern coastal of West Java Province. Data used is ALOS AVNIR-2 and PALSAR which covers research area and have date of acquisition 10 May, 2007 and 5 April 2011 (ALOS AVNIR-2 and PALSAR), and 01 November, 2010 (AVNIR-2) – 03 November, 2010 (PALSAR) for model development, 12 and 21 March, 2011 (ALOS PALSAR) for verification of soil water content, and 5 April 2011 for verification of paddy growing stage. Checking on the field was performed on 5 to 6 June, 2010 and 19 to 20 March, 2011 The ALOS data has spatial resolution of 10 meters (multi spectral) and 25 meters (SAR), as well as <br /> <br /> temporal resolution of 46 days. It can be used to identify the growth period of rice plants and monitor condition of agriculture land in Indonesia. In study area which is located in northern coastal of West Java Province has different types of wetland and give contribution to national paddy production. <br /> <br /> The specific character of paddy can be used to identify a cropping pattern in paddy fields. Analyses were performed using a new algorithm of combination of soil water content derived from satellite data Synthetic Aperture Radar (SAR) and greenness of multi-spectral optical satellite data. Composite analysis of orthogonal transformation that was performed to obtain an algorithm refers to the Landsat Tasseled Cap Transformation (TCT). <br /> <br /> Contribution of this dissertation are (1) method for identifying soil water content and greenness level of paddy plants using remote sensing data, (2) remote <br /> <br /> sensing methods to identify and monitor paddy plant and its growth period, (3) implementation of the COT algorithm was applied in the study area to identify paddy cropping pattern in the region, and (4) implementation of cropping pattern map for estimating area planted and harvested, and paddy production, the need of fertilizer, and irrigation in study area. <br /> <br /> The result of this research shows that soil water content (%) derived from ALOS PALSAR data showed that polarization HH or VV give better result than HV or VH. The COT performed on a new image (PCmultisensor) which is created from multisensor satellite data consisting of NDVI, PCAVNIR2 and 3, and KATPALSAR using PC. Scatter pattern corresponding to greenness as N-Green is PCmultisensor2 with R2 = 0.968, while the band shows soil water content as N-Wet is PCmultisensor3 with R2 = 0.997. PCmultisensor1, as well as PCmultisensor2 on PCmultisensor (NDVI, PCAVNIR2 and 3, KATPALSAR), if associated with PCmultisensor2 (which shows green) has a similar characteristic of scatter pattern of NDVI and rice growth period. Accuracy assessment for soil water content of ALOS PALSAR HH and VV polarization gives an overall accuracy of 72.41% with a kappa coefficient of 0.61. COT algorithm analysis can be obtained maps of rice growing period with overall accuracy of 64.74% and a kappa 0.56. Implementation of this study can be used to monitore agricultural conditions. The analysis of this model is expected to contribute a better management of agricultural land. Therefore it can help farmers (agri-business) and government to plan water needs, as well as an appropriate planting season. Hence, <br /> <br /> it can reduce losses due to crop failure. The results are expected to be used for agricultural planning recommendations.