Thermal Imaging Using Small-Aerial Platforms for Assessment of Crop Water Stress in Humid Subtropical Climates

Authors: Thomson, Steven; ENGLISH, PAT - DREC-MSU; Defauw, Sherri

Submitted to: Asian Conference on Precision Agriculture
Publication Type: Proceedings
Publication Acceptance Date: 6/30/2007
Publication Date: 8/2/2007
Citation: Thomson, S.J., English, P., Defauw, S.L. 2007. Thermal Imaging Using Small-Aerial Platforms for Assessment of Crop Water Stress in Humid Subtropical Climates. In Proceedings of the 2nd Asian Conference on Precision Agriculture,August 1-4, 2007, Pyeongtaek, South Korea. CD-ROM Paper no. DV-01.pdf

Interpretive Summary: As a crop shows signs of water stress, the canopy becomes warmer relative to the surrounding air. In many world regions, there is limited canopy cooling by evaporation due to high prevailing relative humidity. So, detectable differences in temperature between a stressed and non-stressed crop are small. This could be problematic when attempting to use thermal sensing or imagery to determine, for example, when to irrigate. Accurately tracking canopy temperature on a temporal basis requires sensitive and accurate thermal sensing devices and appropriate accounting for differences in prevailing weather. Spatial differences in canopy temperature, however, have been easily illustrated with thermal imagery obtained from aircraft. This study presents a review of past work in both ground-based and aerial thermal sensing with attention paid to limitations in sensing technologies and methodologies for calculating crop water stress indicators. A field study used a thermal imaging camera mounted in an Air Tractor 402B agricultural aircraft to determine spatial differences in cotton (Gossypium hirsutum L.) stress, and relate those differences to soil type differences and crop yield. Geostatistical analysis highlighted areas of the field where assemblages of high crop–soil temperature values were significant. There was high correlation of low yielding zones with areas of the field subjected to the highest canopy temperatures. Ground-truthing along with comparison of color-infrared and thermal image series showed portions of the field where cotton canopy closure was incomplete and the potential for crop heat/water stress was most severe; these areas occurred on clay-rich Tunica soils. Thermal imagery can also be useful for detecting early senescence promoted by these stress factors and could aid in the development of site-specific application of cotton defoliant and harvest aids.

Technical Abstract: Leaf- or canopy-to-air temperature difference (hereafter called CATD) can provide information on crop energy status. Thermal imagery from agricultural aircraft or Unmanned Aerial Vehicles (UAVs) have the potential of providing thermal data for calculation of CATD and visual snapshots that can guide the farm manager to decide where and when irrigation is needed. In addition, thermal imagery offers a streamlined solution to the identification, delineation, and monitoring of field-scale heat contrasts that impact plant productivity over the course of a growing season. A review of past work details research using both ground-based and aerial methods for determining CATD. Limitations of these systems for application of spatial and temporal analysis of pending crop water stress are also presented. A pilot study was conducted to quantify areas of stress and relate these indices to cotton development. Images of an irrigated cotton field were acquired from an Air Tractor AT-402B agricultural aircraft with an Electrophysics PV320T thermal imaging camera mounted inside. For one part of the study, an image obtained five days after irrigation indicated alleviation of crop heat/water stress over close to 95% of the field. A subsequent storm event resulted in 37.8 mm (1.49 in) of rainfall; a thermal image acquired several days later charted the subsurface redistribution of soil water and the general lack of crop water/heat stress across most of the field (with the exception of two prominent clay-fingers, characterized by incomplete canopy closure). The thermal imagery time-series was composited to produce a cumulative thermal image that spanned close to five weeks of elapsed time. Geostatistical analysis (i.e., univariate analysis using Local Moran’s I Spatial Autocorrelation or LISA) highlighted areas of the field where assemblages of high crop-soil temperature values were significant. The overall pattern of significant spatial relationships for this particular cotton field demonstrated the rather tightly-coupled linkages of low yield zones with areas of the field subjected to the highest temperatures as well as the pairing of high yield zones with cooler canopy temperatures. Ground-truthing coupled with the thermal image from later in the season (31 August 2006) indicated that most cotton plants in the clay-rich portions of the field had over 50% open boll. Even though CATDs could not be directly assessed in this pilot study and subsequently used to develop a Crop Water Stress Index (CWSI), the patterns of thermal zonation developed in this highly heterogeneous field revealed areas where crop water and/or heat stress repeatedly occurred and yield was negatively impacted. Thus, thermal imaging should be a useful alternative tool in the identification of early senescence promoted by water/heat stress in highly heterogeneous cotton fields, and also be of aid in the development of site-specific application of defoliants/harvest aids.