Gy. Tarcsai

Crop yield estimation by satellite remote sensing

Two methods for estimating the yield of different crops in Hungary from satellite remote sensing data are presented. The steps of preprocessing the remote sensing data (for geometric, radiometric, atmospheric and cloud scattering correction) are described. In the first method developed for field level estimation, reference crop fields were selected by using Landsat Thematic Mapper (TM) data for classification. A new vegetation index (General Yield Unified Reference Index (GYURI)) was deduced using a fitted double-Gaussian curve to the National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) data during the vegetation period. The correlation between GYURI and the field level yield data for corn for three years was R 2=0.75. The county-average yield data showed higher correlation (R 2=0.93). A significant distortion from the model gave information of the possible stress of the field. The second method presented uses only NOAA AVHRR and officially reported county-level yield data. The county-level yield data and the deduced vegetation index, GYURRI, were investigated for eight different crops for eight years. The obtained correlation was high (R 2=84.6–87.2). The developed robust method proved to be stable and accurate for operational use for county-, region- and country-level yield estimation. The method is simple and inexpensive for application in developing countries, too.

Routine whistler analysis by means of accurate curve fitting

Abstract A new method has been developed for the reliable, routine analysis of whistlers with unknown nose frequency and causative sferic. The method is based on the least squares estimation of two parameters (D0 and fHE) in Bernards (1973) approximation for the dispersion, and of a third parameter locating the causative sferic. A comparison on measured whistlers showed that the new method has certain advantages over the method described by Rycroft and Mathur (1973) and is more reliable than the procedure developed by Ho and Bernard (1973). The equatorial electron density ne and electron tube content NT values were computed using Parks (1972) expressions. A careful error analysis performed on about one hundred whistlers (1.4 ≦ L ≦ 3.3), analysed by the new method, indicated that errors in ne and NT are less at about L = 3.1–3.3 and L = 2.3–2.5, respectively. The analysis of travel time residuals proved to be useful in studying the measurement errors and the approximations that have been made.

Average electron density profiles in the plasmasphere between L = 1.4 and 3.2 deduced from whistlers

Abstract 985 whistlers recorded at Tihany, Hungary (L = 1.9) between December 1970 and May 1975 have been processed for equatorial crossing radius L, equatorial electron density neq and tube electron content NT. The obtained L values lie in the range L = 1.4-3.2. The median values of neq and NT as a function of L fit well to the density profiles published by Park et al. The results also indicate that at lower L values the whistlers are ducted by stronger density enhancements.

Yield estimation for corn and wheat in the Hungarian great plain using Landsat MSS data

Abstract Models for the forecasting of crop yields using remotely-sensed satellite data are studied intensively worldwide. After reviewing the experience gained by other researchers in this field, we selected procedures which might be suitable for the estimation of corn and wheat yields in Hungary. In order to study the relations between various remotely-sensed spectral data (and their combinations) and the actually measured final yields we investigated archived agricultural and Landsat MSS spectral data for 1984. A linear relation has been sought and found between the yield data for 47 corn and 55 wheat fields in Hajdu-Bihar county and various weighted and summed spectral quantities. Among the vegetation indices derived from satellite data and corrected for atmospheric effects the most promising were the spectral indices sensitive to the green biomass. The latter, summed over a certain period in the growing season, exhibited a regression of 45-86 per cent, depending on the crop and the period of summatio...

Fine structure of whistlers recorded digitally at Halley, Antarctica

Abstract Two whistlers recorded digitally at Halley, Antarctica (L = 4.3) were analyzed by matched filtering with 10 Hz frequency resolution. For the construction of the matched filter, a more realistic description of the whistler waveform than has been used for lower latitude whistlers was applied. The ( ƒ, t ) pairs obtained, together with the corresponding magnitudes, gave a high resolution dynamic spectrum which revealed the fine structure of the whistlers. Whistlers which appeared as well defined, discrete, single traces on a conventional spectrogram, turned out to be composed of several components covering various parts of the spectrum. Further analysis of the strongest and longest component of a whistler resulted in high resolution travel time residual curves similar to those obtained by the ray tracing method or by averaging of numerous occurrences of whistlers in the same duct. The matched filter method of analysis can provide new insights into the details of ducted whistler propagation. It also yields the amplitude variation vs time (or vs frequency) along the whistler trace, thus potentially giving information about the spectrum of the source spheric, and/or the frequency dependence of magnetospheric amplification due to wave-particle interactions.

Trace splitting of whistlers: A signature of fine structure or mode splitting in magnetospheric ducts?

Previously, we reported on the discovery of fine structure in whistler data received on the ground at Halley, Antarctica. This structure was not apparent in conventional spectral analysis but was revealed by the technique of digital matched filtering. We have now examined a larger data set, and a commonly observed phenomenon is that single whistler traces become split into two, over various frequency ranges. Examples are presented in the form of time-transformed spectrograms in which reference model whistlers are represented as vertical lines. The splitting is typically 5–15 ms (about 0.5% of the total whistler travel time) and extends over frequency ranges of a few hundred hertz which may occur anywhere between the upper and lower cutoff frequencies of the whistler. The splitting may be either symmetrical or unsymmetrical with respect to the unsplit trace. The effect is unlikely to arise in the spectrum of the lightning source or from propagation under or through the ionosphere. It may, however, be a signature of field-aligned fine spatial structure in plasmaspheric density, and hence refractive index, in the whistler duct. For simple longitudinal propagation, electron density fluctuations of the order of 1% and spatial scale sizes of the order of 50 km in the equatorial plane are implied. It seems possible that the observations could also be interpreted in terms of the mode theory of ducted propagation, assuming the excitation of two modes with group velocities differing by a few tenths of a percent.

Error sources and travel time residuals in plasmaspheric whistler interpretation

Abstract In the interpretation of observed whistlers by curve fitting, systematic travel time residuals appeared which were studied by extensive simulations using ray-tracing, numerical integration and curve fitting. The residuals were found to originate from the commonly used approximations in the refractive index and ray path of whistler mode waves, which result in travel time increments or decrements, not accounted for in whistler interpretation. These approximations and the assumed form of the electron density distribution also lead to systematic errors in the diagnostics of plasmaspheric electron density by whistlers. In addition, the effects of other error sources, including random measurement errors, are also reviewed briefly. It is shown that the fine structure of residual curves is connected to propagation conditions. Thus, their study may yield a new research tool for studying whistler trapping, ducting structures and other features of whistler propagation. The application of residual analysis in conjunction with digital matched filtering of whistlers seems to be especially promising for further whistler studies.

Ionosphere-plasmasphere electron fluxes at middle latitudes obtained from whistlers

985 whistlers observed between 1970 and 1975 in Hungary have been processed for equatorial plasmaspheric electron density and tube electron content above 1000 km (NT). The hourly median value of the tube electron content exhibits a diurnal variation with an amplitude of 1 x 10 to the 13th electrons/sq cm-tube. 75 percent of the electron flux values obtained from the time variation of the tube electron content are lower than 6 x 10 to the 8th el per sq cm s, while in some cases the fluxes reach a value as high as 3 x 10 to the 9th el per sq cm s. Between 17 and 04 LT the dominant flux direction is toward the ionosphere. The data also indicate that the day to day filling of the plasmasphere after magnetic disturbances continues through several days without exhibiting saturation, with higher filling rates for lower values of average Kp.

Whistler inversion by spectral expansion

Abstract A method has been developed for the inversion of the integral equation connecting whistler propagation times with the distribution of electrons along the propagation path. The path is taken to be a geomagnetic dipole field line, but the method assumes no model for the field aligned distribution of electrons. The calculation is based on the Backus-Gilbert technique ( gilbert , 1971; parker , 1977) and represents the electron density distribution as a sum of orthogonal functions, the coefficients of which are computed from the time delays measured on the whistler trace at a series of frequencies. The application of this method to synthetic and observed whistlers has demonstrated the feasibility of obtaining magnetosphere electron densities. At the same time it is found that some of the approximations of conventional whistler analysis (omission of ionic effects and a term of +1 from the phase refractive index) are inadequate and lead to certain difficulties in the inversion of higher latitude ( L ≥3) measured whistlers.

Correction of atmospheric effects of satellite remote sensing data (Landsat MSS-NOAA AVHRR) for surface canopy investigations

Abstract A new correction method for atmospheric effects in Landsat-MSS and NOAA AVHRR data is presented which uses only the remotely-sensed multispectral data. The method is based on a new quasi-single-variable radiative transfer model, and as a first step we assumed that the surface is covered by vegetation. For Landsat-MSS data the method was developed for the tasseled cap indices using known empirical relationships among them. For NOAA AVHRR data ‘ cap-like’ indices and the average reflectance of the average canopy in the visible band known from Landsat-MSS data were used. The method was used in yield forecasting project in north-eastern part of Hungary and provided a significant enhancement in the quality of remotely sensed data.