·Home ·Table of Contents ·Workshop - Landmine Detection Equipment

Buried mine and soil temperature prediction by numerical model P. Pregowski, PIRS, Pregowski InfraRed Services, Warsaw, PL, Email : pirspp@warman.com.pl W. Swiderski, Military Institute of Armament Technology, Zielonka, PL, Email : zak_13@witu.mil.pl R.T. Walczak, K. Lamorski, Institute of Agrophysics, Polish Academy of Sciences, Lublin, PL, Email : rwalczak@demeter.ipan.lublin.pl Contact

Abstract

This paper presents elaborated numerical models and also shows thermal signal and noises dependencies on such parameters as: time and space variability of moisture and density of soil, buried mine and soil features, as well as chosen environmental conditions. Numerical experiments and measurements, that were carried out, clarified some of physical limitations for applying simple procedures of IR detection.
Keywords: Thermal imagery, buried mines, heat transfer, soil temperature, numerical simulation

1. INTRODUCTION

In recent years, many world-wide institutions have started the work to improve mine disposal effectiveness. Among about 20 technologies being presently developed, IR imaging is one of essential ones- although strong dependence of results on both measuring conditions and operator skills, is a weakness of this method. Thermal detection can be effective only when high enough difference in radiant signal exists at the surface above the mine, compared to region free of this target. Due to very high thermal inertia of the soil, usually it is assumed, that underground mines manifest themselves only as result of changes in solar activity. Because of that, the ability to predict proper time and detection procedure, plays particularly important role. A few years ago we began collecting experimental data on the field stand-ups with real mines[1]. Our assumptions were, that we would be able to find correlation, required for planning the missions - which failed. Much deeper understanding of the buried mine's signature, proved to be indispensable. To describe mine's image in the output of thermal imager, a few different models have to be combined. Models to describe the influence of the IR camera features and models to predict the radiometric aspects of incoming signals, almost similar for land and buried mines, appeared to be relatively well developed. Study of the physics-related papers[2-5] showed considered, here, problem as not developed enough. It is probably due to high complexity and variability of the soil physics processes[6]. Models presented here were elaborated as the base to the next ones, more specialised and simplified, through links with various bases of the reference data. More details about these models and experiments had been presented by us earlier[7].

2. PHYSICS AND MODELLING OF THE MINE AND SOIL TEMPERATURE

Detected signals depend on the surface temperature and emissivity distributions. The differences in emissivity, caused by process of burying the mine due to soil shrinking-swelling cyclic processes, usually decreases quickly with time, so their influence had been omitted by us. For modelling, the buried mines temperature component is the most essential. Its distributions on the soil surface strongly depend on the state of processes of mass and energy exchange (radiation and convection, evaporation and water condensation, supply of water through precipitation and gaseous exchange), both directly connected with the temperature profile of the soil. This profile depends on the next variables, as e.g. dynamic nature of heating-cooling conditions, which decides about the mass and heat flow in soil; the time and space variability of soil moisture and air content; soil thermal properties, which are result of the natural processes of soil formation, under specific topographic conditions. The role of these factors differ in relation to the structure of the soil's upper layer, formed by biological life and human activity, as e.g. mine - burying. Fig. 1 presents amplitude-phase component the soil temperature changes versus cyclic diurnal variation of solar activity, for two homogeneous soils at different depths.

Fig 1 :

Formula applied above, proposed by Wijk and de Vries, is too general to solve considered problems.
The method of our modelling can to be shortly characterised as follows. We assume that there is no natural spatial heterogeneity of soil thermal parameters, and that the heat and water fluxes depend on the atmospheric conditions, following to the formula : R_n+H_s+LE+G=0 where: R_n - net radiation at the soil surface, H_s - sensible heat flux exchanged between soil surface and atmosphere, L - latent heat of vaporizing, E - water flux at the soil surface, G - heat flux at the soil surface.
The following processes were taken into account: water conductivity resulting from soil water potential gradients; thermal conductivity resulting from temperature gradients; surface soil processes: vaporising and energy exchanging with the environment; variability of thermal properties caused by mine buried in soil, and by differences in water content and soil density. Flux of energy in the soil is calculated from the Fourier's equation using the of average value of thermal conductivity coefficient "l " calculated on the base of model elaborated in Lublin Institute of Agrophysics[8]. Transport of water in the soil profile is calculated using Richard's equation and water retention in the soil using Mualem-Van Genuchten model. Mine is to be considered as a roller, so the whole system, soil-buried mine, has axial symmetry, what simplified the calculations a lot. Differential equations, which describe heat and water transport in the soil, are spatially approximated using finite differences method on a rectangular grid. For these approximation for the space net, Crank-Nicholson's method was partly modified.

2. RESULTS OF SIMULATIONS AND EXPERIMENTS

Modelling of all the basic soil thermal characteristics in dependence of soil humidity, soil density, soil temperature and mineral soil composition, allowed us to prove strong non-linear influence of the moisture for thermal conductivity and diffusivity of soil. Fig. 2 allows to deduce, that differences in space distribution of moisture and structure of the soil, could induce temperature signals analogous to those induced by buried mines, named false alarms.Fig. 3 presents comparison of simulated and measured data about thermal profile of sandy soil, disturbed by the plastic mine located 10 cm to this profile and at depth of 5cm.

Fig 2 :

Fig 3 :

Fig. 4 presents thermal visibility of the mines buried at depth of 5cm, - resulted from various amplitudes of the solar irradiation(200-500)Wm^-2 and unchanged cyclic changes the air temperature and the solar heating versus time. Inversions of thermal signatures and their delay in time, caused by both differences between solar activity and air temperature maximums as well as top layer thickness influence, are clearly seen. Fig. 5 shows simulated temperature distributions in soil at the end of heating (a) and after long cooling (b). Fig. 6 is simulation of thermal signature of the buried mine at the end of heating.

Fig 4 :

Fig 5 a,b:

Fig 6 :

Fig 7: collects example of IR thermograms of plastic mine at various time of heating/cooling process.

Thermograms as above were obtained at laboratory, under controlled conditions, with the aim to verify and to tune the numerical model. Similar tests in the field, altough much more complicated and time consuming, are planned for this summer and autumn and will be executed on the specialised agro-climatical stand in IA PAS Lublin.

3. CONCLUSIONS

The elaborated model enabled us to use simulation prediction as the basis for experiments, which otherwise would be difficult or impossible to perform. Obtained results in quantitative form, explain the causes of our troubles in determining what is what and why just the particular spot marks the mine - proving a few weak sides of the thermographic method. Even weak differences in soil moisture and/or density distributions can act as sources of false alarms. These relationships occur particularly strong for low moisture and high density of soil. The importance of these effects decreases when diurnal amplitude of temperature changes increases. To increase the model's accuracy in representation of the real-world, some modernisation, as e.g. more detailed description of the surface, feature influences and conductivity of water vaporizing or energy changes caused by water vaporising and condensation of vapour in the soil are planned. In the present state, elaborated models are unsuitable for the field application.

Acknowledgements

We would like to recognise help provided by B. Usowicz from Institute f Agrophysics and D. Szabra from MIAT. Research was supported by the State Committee for Scientific Research under biennial project No 0-T00A-015-14, finished at 1999.

References

1. P. Pregowski, W. Swiderski, - "Detection of minefields using IR sensing and time-domain treatment method", Proc. of Conf. Detection Technologies for Mines and Minelike Targets, SPIE Vol. 3079, p.791-800, (1997).
2. J-R. Simard, "Improved Landmine Detection Capability (ILDC): Systematic approach to the detection of buried mines using passive IR imaging", Proc. of Conf. Detection Technologies for Mines and Minelike Targets, SPIE, Vol.2765, p.489-500, (1996).
3. K. Russel, J. McFee, W. Sirovyak: "Remote performance prediction for IR imaging of buried mines", Proc. of Conf. Detection Technologies for Mines and Minelike Targets, SPIE, Vol.3079, p.762-769, (1997).
4. Y.H.L. Janssen, A.N. de Jong, H. Winkel, F.J.M. van Putten: "Detection of surface laid and buried mines with IR and CCD cameras, an evaluation based on measurements", Proc. of Conf. Detection Technologies for Mines and Minelike Targets, SPIE, Vol.2765, p.448-459, (1996).
5. G. Maksymonko, B. Ware, D. Poole: "A characterization of diurnal and environmental effects on mines and factors influencing the performance of mine detecting ATR algorithms", Proc. of Conf. Detection Technologies for Mines and Minelike Targets, SPIE, Vol.2496, p.140-151, (1995).
6. P.Pregowski, W Swiderski "Application of IR thermography for detection of the buried mines - prediction and procedures", Proc. of MATEST'99, Cavtat, Oct.,p227-232, (1999).
7. P.Pregowski, W Swiderski, W.T. Walczak, K. Lamorski, "Buried mine and soil temperature prediction by numerical model", Proc. of Conf. Detection Technologies for Mines and Minelike Targets, SPIE AeroSense Conf., Orlando, 25-27 April, 2000 (in the process)
8. B. Usowicz : "Statistical-Physical Model of Thermal Conductivity in Soil.", Pol. J. Soil. Sci., 25,p. 25-34, (1992).

© AIPnD , created by NDT.net

|Home|    |Top|