Evaluation of water stress on a crop using the portable hyper spectral imager
Hidetomo FUJIWARA, Takahiro ENDO, Yoshifumi YASUOKA
Institute of Industrial Science
University of Tokyo
4-6-1 Komaba, Meguro-ku, Tokyo, Japan, 153-8505
Tel: +81-3-5452-6415 FAX : +81-3-5452-6410
E-mail: hidetomo@iis.u-tokyo.ac.jp
Abstract We measured changes of spectral characteristics from crops water stress in different
scales of leaf and plants by using the portable hyper spectral imager which had high spatial and
spectral resolution. Our research object is to evaluate of water stress on the basis of experiments.
The vegetation of our experiment is soybeans. Our plantation crop had been divided into two
groups, one with enough water for growth and the other under water stress. Measurements of
soil water, water potential and SPAD value has been done to relate these parameters. From the
measurements, it is found that, spectral sensitivity due to water stress had been found be highest
in the near infrared region as high correlation was found between SPAD value and spectral
change in this region.
1. Introduction
Crop failure due to water stress is an important problem needing adequate research attention as
it has been found that the sixty percent of crop damages is from water stress. Remote sensing
has been found to have good potential in mapping water stress from for better managerial
decision making. With its synoptic and regular temporal capability, we think remote sensing can
be quite effective in early detection and forecasting of crops failure. Past researches about water
stress are concentrated mostly on a certain scale and doesn’t relate the results in different scales
of leaf, plants and vegetation in a forward manner. In this research, we measured changes of
spectral characteristics from crops water stress in different scales of leaf, plants and vegetations
by using the portable hyper spectral imager which had high spatial and spectral resolution. Our
research object is to evaluate of water stress on the basis of experiments. The vegetation of our
experiment is soybeans as it is one of the demanded crops in the world. Our plantation crop has
been divided into two groups, one with enough water for growth and the other under water
stress. Measurements of soil water, water potential and SPAD value’ which is spectral signal
from the vegetation’ has been done to relate these parameters. From the measurements, it is
found that, spectral sensitivity due to water stress has been found be highest in the near infrared
region as high correlation is found between SPAD value ‘direct measurement’ and spectral
change ‘indirect measurement’ in this region.
