Monitoring of maize damage caused by western corn rootworm by remote sensing
Gizella Nádor(1), Diána Fényes(1), László Vasas(2) and György Surek(3)
(1) Institute of Geodesy, Cartography and Remote Sensi, Bosnyák tér 5, H-1149 Budapest, Hungary
(2) Agricultiral Office of County Békés, Szarvasi út 79/1, H-5600 Békéscsaba, Hungary
(3) MLOG Ltd., Klebersberg kunó u. 36, H-1158 Budapest, Hungary
As one of the consequences of global warming, the gradual dispersion of western corn rootworm (WCR) is becoming a serious maize pest in Europe and all over the world. The WCR was introduced to Europe from the USA. First it was detected in Europe near Beograd in 1992. The WCR has spread from its initial infestation point to a range of several hundred kilometres, affecting many countries in the region including Hungary first in 1995.
The structure of a healthy, WCR free maize field shows straight rows in a clear order and upstanding maize stalks. The WCR infection causes wilted broken corn-stalks lying randomly on the ground. That means the damage itself causes physical and visible disorder in the maize field.
According to our experience, to apply remote sensing in a proper way, the key is to have the most possible detailed and versatile description based on ground assessment of the differences between the healthy and the WCR damaged fields. Our goal is to assess and identify the disorder and structural changes caused by WCR using polarimetic radar images (ALOS PALSAR).
We used 3 different individual features (Mono-maize feature, Optical feature, Radar feature) derived from remote sensing data to identify larval damage. These are as follows:
• The result of monoculture maize field assessment (Mono-maize feature)
• Vegetation index derived from optical satellite data decreases if there is larval damage (Optical feature)
• Identification of the disorder and structural changes caused by WCR using polarimetic radar images (Radar feature)
The integrated assessment of the 3 features can give more accurate WCR damage identification than to assess features separately one-by-one.
The accuracy of damage identification was 57-70 % with the one-by-one evaluation of the remote sensing features. The monoculture feature gave the most incorrect results. The other two features presented about the same results, did not show significant differences. It clearly proves that considering the theory we made a good decision about using radar data to identify WCR damage. The important outcome of our mission is that polarimetric radar data can add a competitive advantage in remote sensing vegetation assessments and detecting structural changes.
The accuracy of damage identification is higher than 80% with the integrated evaluation of the 3 features. This result is 10 % better than with the one-by-one evaluation. This project demonstrates the potential in the integrated assessment of optical and radar satellite images to assess and identify the disorder and structural changes caused by WCR larvae as well as other agricultural damage.
Based on the first results, introduction of polarimetric radar technique significantly increased the accuracy. The further development of this technology and using more radar images will create the possibility to accomplish a more accurate damage identification system. By the successful development, this project can effectively contribute to the WCR identification, spread-monitoring and control in Hungary as well as in the European Union. The amount of information can be used efficiently in plant protection; moreover can be useful for farmers, pesticide producers, state authorities and research institutes. The objective of this project is unique in the European Union.
The project was carried out by the support of the Hungarian Space Office and Ministry of Environment and Water. The ALOS PALSAR data were provided by ESA (ESA EO CAT-1 5162) and FÖMI. The reference data were collected by the experts of the Central Agricultural Office of Békés County.