Analyzing rice farming between sowing and harvest time with Sentinel-1 SAR data


Creative Commons License

Polat A. B., Balık Şanlı F., Akçay Ö.

Advanced Remote Sensing Journal (ARSEJ), vol.2, no.1, pp.34-39, 2022 (Peer-Reviewed Journal)

  • Publication Type: Article / Article
  • Volume: 2 Issue: 1
  • Publication Date: 2022
  • Journal Name: Advanced Remote Sensing Journal (ARSEJ)
  • Page Numbers: pp.34-39
  • Çanakkale Onsekiz Mart University Affiliated: Yes

Abstract

Agriculture has always been in an important position throughout human history. Today, the  development  of  technology  has  accelerated  studies  to  increase  productivity  in agriculture. With the use of remote sensing in agriculture, different crop types in large regions could be observed and their differences from each other could be examined with a  spectral  sight.  With  the  observations  obtained,  instant  surface  monitoring  in  the agricultural sector makes it possible to perform analyzes. In the study, the paddy fields, where the rice product was named at the time of first planting, were examined by remote sensing  method.  Differences  in  Synthetic  Aperture  Radar  (SAR)  observations  were analyzed between the first crop sowing and harvest time. In addition, in order to check the  consistency  of  the  results,  the  differences  in  the  values  obtained  according  to  the representation of the samples distributed in the field were determined. Considering the results, it was seen that the lowest backscatter values were obtained for the paddy fields in  the  35-day  period  after  the  first  plantingtime  and  these  values  increased  as  the harvest  time  approached.  There  is  an  approximately  69%  change  in  the  lowest  and highest  mean  backscatter  values.  Finally,  when  the  time  series  analysis  is  performed according  to  the  control  samples  in  the  field,  it  has  been  determined  that  the  points represented by a single pixel have a more irregular distribution comparing the samples obtained   in   the   form   of   polygons.   This   shows   that   pixels   cannot   be   evaluated independently due to noise in SAR data.