Open AccessCurrently, a promising trend in remote sensing of environment is to monitor the vegetative cover: evaluate the productivity of agricultural crops; evaluate the moisture content of soils and the state of ecosystems; provide mapping the sites of bogging, desertification, drought, etc.; control the phases of vegetation of crops, etc.
Development of monitoring systems for remote detection of vegetation sites being under unfavorable conditions (low or high temperature, excess or lack of water, soil salinity, disease, etc.) is of relevance. Optical methods are the most effective for this task. These methods are based on the physical features of reflection spectra in the visible and near infrared spectral range for vegetation under unfavorable conditions and vegetation under normal conditions.
One of the options of optoelectronic equipment for monitoring vegetation condition is laser equipment that allows remote sensing of vegetation from the aircraft and mapping of vegetation sites with abnormal (inactive periods of vegetation) reflection spectra with a high degree of spatial resolution.
The paper deals with development of a promising dual-spectrum method for laser remote sensing of vegetation. Using the experimentally measured reflection spectra of different vegetation types, mathematical modeling of probability for appropriate detection and false alarms to solve a problem of detecting the vegetation under unfavorable conditions (with abnormal reflection spectra) is performed based on the results of dual-spectrum measurements of the reflection coefficient.
In mathematical modeling, the lidar system was supposed to provide sensing at wavelengths of 0.532 μm and 0.85 μm. The noise of the measurement was supposed to be normal with zero mean value and mean-square value of 1% -10%.
It is shown that the method of laser sensing of vegetation condition based on the results of dual-spectrum measurement of the reflection coefficient at wavelengths of 0.532 μm and 0.85 μm makes it possible, even with a measurement error of 10%, to detect vegetation sites that are under unfavorable conditions, with the probability of appropriate detection close to 1 and the probability of false alarms close to 0.