CA1103790A

CA1103790A - Ceilometric method and apparatus

Info

Publication number: CA1103790A
Application number: CA305,490A
Authority: CA
Inventors: Frank Frungel
Original assignee: Individual
Current assignee: Individual
Priority date: 1977-06-15
Filing date: 1978-06-14
Publication date: 1981-06-23
Also published as: FR2394816B3; DE2726999C2; GB2000411A; JPS6114470B2; FR2394816A1; DE2726999A1; JPS547368A

Abstract

ABSTRACT OF THE DISCLOSURE
The total range to be explored consists of a suc-cession of range-steps. For each range-step, a transmitter emits upwards at least one laser pulse, which is reflected downward by clouds onto a gated receiver. The gating interval of the re-ceiver, relative to the emission instant, determines what range-step is being explored. The gating interval is progressively shifted, to run through the range-steps. For each range-step, reflected laser pulses received during the gating intervals, if any, are integrated. Because the power attenuation of the re-flected pulses is an approximately quadratic function of height, the relationship between the range-step value and the integrated signal is made approximately quadratic. In the lower part of the total range, this is done by progressively overlapping the conical beam patterns of transmitter and receiver, to create an overlap which increases quadratically with height. At least in the re-maining upper part, an integration schedule is established, fixing for each range-step the number of pulses integrations performed before going onto the next range-step. The number of integrations performed per range step increases approximately quadratically as a function of range-step value or height. The beam-overlap ex-pedient is used because of its simplicity, but is limited as to range. The integration-scheduling expedient is not limited as to range but cannot always be devised to span the transition between lower and upper parts of the total range. The two expedients in conjunction make it possible to span across the middle part of a large ceilometric range.

Description

1 The present invention relates to ceilometric methods and apparatus, operating on the shifted gating inter~al versl~s travel-time basis. In this type of system, high-power laser pulses are emitted by a transmitter vertically upwards into -the volume of air whose ceiling is to be ascertained. The thusly emitted pulses are reflected downwards by clouds, haze and the like and received by a receiver which adjoins the transmitter, e.g., mounted on the ground. The ceiling, i.e., height from ground level to clouds above, is ascertained on the basis of the travel time of the thusly transmitted and reflected laser pulses.
However, the travel-time measurement is not performed directly, but instead indirectly by periodically shifting with respect to time the sampling interval of the receiver. Furthermore, the measurements are performed repetitively or redundantly and the reflected pulses are integrated, to yenerate signals or sufficient j magnitude for si~nal-processing purposes.
An example will make this clearer. Assume that the ceilometric system is to be capable of detecting ceilings ranging from 3000 ft. up to 5000 ft., and is to have a resolution of 50 ft. First, the system tries to ascertain whether the ceiling is 3000 ft.; then it tries to ascertain whether the ceiliny is 3050 ft.; then whether it is 3100 ft.; etcO; going up in 50-ftu steps at periodic intervals. For e~ample, during the first measuring step, i.e., when the system tries to ascertain whether the ceiling is at 3000 ft., the system operates as follows: First, the laser-pulse transmitter emits a laser pulse vertically upwards. The cooperating reflected-pulse receiver of the system operates on a gated basis. It is enabled for pulse reception only during the time interval during which the thusly emitted laser pulse would be received at the receiver, if indeed the ceiling were 3000 ft. A

2 ~ r 7~

1 sampling gate in the receiver system is synchronized to -the transmitter, and a programmed time-delay stage responds to the triggering of the pulse transmitter by gating the receiver at just -the time that the pulse would be returned, for a ceiling of 3000 ft. I~, during this gating interval a-t the receiver, no reflec-ted laser pulse is received, it follows that the ceiling is not at 3000 ft.
Therefore, -the system next tries to ascertain whether the ceiling is at 3050 ft~ A laser pulse is again emitted, but the gating interval of the receiving system is shifted, i.e., occurs a little later now rela-tive to the triggering of the laser-pulse transmitter "opening" the receiver at the time interval which would correspond to recep-tion of the transmitted laser pulse if it were reflected o~f a 3050-ft. cloud. And so it goes, the gating interval of the receiver being shifted, step by step, by time intervals corresponding to 50-ft. ceiling intervals.
However, because of the level of power employed for the laser pulses, during each of the aforementioned 50-ft. range steps, the measuring process is repeated, e.g., 100 or 1000 times, before going on to the next 50-ft. range step, i.e., before shifting the gating interval for the next range step. During one such range step, e.g., 100 or 1000 pulses are emitted. In the event the range step at which the system is presently working really does correspond to the ceiling, reflected pulses are re-ceived at the receiver during the gating interval for this range step, e.g., 100 or 1000 such pulses. These received signals, all corresponding to one range step, are integra-ted, to generate a signal whose magnitude is large enough for signal processing.
Thus, the ceilometric system not only increases -the range on a step-by-step basis but, within each range-step, per-

3~

1 forms a gated-interval transmission and attempted reception of a reflected lasex pulse a rather high number of times, the reflected energy received (if any) during this many times repeated one-range~step measurement being integrated, to make sure that the reflected signal received (if any) has a chance to build up to a magnitude high eno~lgh for accurate detection and subsequent signal processing. Only then, does the system shift its gating interval for the next-higher range step.
In the earlier years of laser technology, the po-tential of laser radiation for damage to human eyes was not ap preciated. Accordingly, without inhibition, it was a simple matter to use transmitted laser pulses of very high peak power in so-called lidar ceilometers, to measure ceiling, i.e., cloud height. Even then, however, dif~iculties were encountered with respect to service life of the instruments; in the case of un-manned weather stations and the like, it is of course very im-portant that the system operate reliably for months without in-spection, maintenance or service work.
Simple ceilometric systems operating on a travel-time basis presented an immediate additional disadvantage relative to conventional ceilometers, the latter operating on a geometrical basis, similar to theodolites, and being operative for measuring the angle included between the ground and radiation reflected from a cloud. Specifically, when the transmitter and the receiver closely adjoined each other, or were indeed even coaxially ar-ranged, the angle between emitted pulsed radiation and received reflected radiation was most extremely small. The intensity of back-scattering of radiation from atmospheric aerosols, however, is approximately inversely proportional to the 4th power of this angle. Accordingly, with traveltime systems of this type, intense reflections of radiation were received off of even thin smog ancl -- 4 ~

l haze layers, and even off of air inversion layer boundaries, -the radiation transmitted and received by a system operating on the conventional angle-measurement basis being ~ery large, due to the large length of the measuring base. As a result, lidar ceilo-metric installations have led to a great variety of ceilometric systems which, however, especially in airports for flight-safety use, do not find very comprehensive application. For example, very often a low-lying cloud is registered, leading to a prohi-bition against landing by approaching aircraft, whereas the cause of the registration is merely smog and haze, through which the pilots of such aircraft can see quite clearly, they accordingly not understanding why they are being prohibited from attempting a landing.
Another problem with such ceilometric systems relates to attempts to deal with international eye-safety codes concerning the emission of pulsed laser energy for such purposes. The average emitted laser power must ~e kep-t within prescribed limits. This has led to the development of certain prior-art techniques, devised to meet such co~es while still being able to perform the requisite ~0 ceilometric measurements. It is more appropriate to discuss these in detail further below. The net effect of these techniques is to lower the average emitted laser radiation in a way which neces-sarily increases the time required for a complete runthrough through the range-steps constituting the ceiling range to be investigated. Unfortunately, other international codes and regu lations, particularly concerning flight-safety systems at air-ports/ place limits on the amount of time permitted for such measurements, e.g., by requiring that four Eresh such measurements be furnished per minute, leaving only 15 seconds ~ithin which to somehow run through the range of interest. Of course~ this type 1 of regulation does not apply to cellometric systems used ~or metereological applications, e.g., in isolated unmanned weather stations. Indeed, the ceiling ranges of interest for airport flight-safety systems are on the low side, whereas those of purely metereological interest tend to extend up much higher. ~ttempts to devise ceilometric systems which meet the average-emitted laser power and measurement-rate standards of flight-safety systems, besides facing the difficulties brought about by the conflicting nature of those two safety considerations, tend to result in systems which are better suited for low-level measurements, and therefore not well suited for metereological purposes. Thus, flight-safety needs conflict with each other, and their compro-mises often conflict with metereological needs, making it ex-tremely difficul-t to provide a general-purpose ceilometric system which can be used alternatively or simultaneously for flight-safety and metereological purposes. These problems are discussed in somewhat greater detail below.
It is a general object of the invention to provide a ceilometric system which is, on the one hand, of comparatively very simple and reliable construction and operation so as to be well suited for unmanned weather-station applications and whose operation is suited to such applications, while at the same time being suitable for airport flight-safety applications.
Accordingly, it is an object of the invention to provide a ceilometric system which works in a way making it possible to keep average emitted laser radiation low, while not increasing the times required for measurement runthroughs beyond those al-lotted by international airport-safety-system codes, without sacrificing system accuracy and effectiveness at either the low or high ends of the total ceilometric range which would be of interest a ~7~

1 when both airport and metereological applications are to be comprehended. Furthermore, this is, above all, to be done in the simplest possible way.
According to a broad concept of the invention, this is to be accomplished by using a certain combination of techniques.
These techniques relate to the problem, discussed in greater detail below, of rnatching the signal strength of the system to the range-step program of the system. For example, in the prior art, i~ is g~nerally considered necessary that the gain of the amplifiers of the system be altered as the square of the range, so that the signal for any given range step will be more or less normalized.
The present invention can achieve this kind of result in a different way, devised to achieve the goals set forth earlier. For the part of the ceilometric range from zero up to about 1000-2000 feet, for example, corresponding to a laser-pulse travel time of 2-4 micro-seconds, the transmitter-to-receiver transmission factor of the system is progressively increased, approximately as the square of range, to normalize signal strength, by a pllrely geometric technique.
In particular, the conical beam patterns of the transmitter and receiver of the system are made slightly convergent, to produce a beam-pattern overlap whose size increases approxima~ely as the square of range through this part of the ceilometric range.
Furthermore, for range-steps beginning at the upper end of this lower range (i.eO, beginning at 1000 or 2000 feet) or beginning even lower, the number of received reflected laser pulses integrated for each range-step, used to perform the measure-ment for that range-step, is progressively increased in accordance with a predetermined schedule, as one proceeds up towards the upper limit of the system's range, e.g., 10,000 feet. Thus, for the lower range-steps in this part of the range, the system may 1 for example take in only one received reflected laser pulse, then for higher range-steps a few pulses per range-step/ and so on, with several 100 pulses per range-step being taken for range-steps at the upper limit of the range of the system, preferably with no progressive adjustment of the gain of the receiver of the system, and pre~erably with no adjustment in the per-pulse energy of the transmitter.
In this way, the desired normalization of signal strength throughout the sequence of range-steps constituting the range can be performed with very primitive basic operating circuits, i.e., not gain-controlled or per-pulse power-controlled. Instead, the normalization is in part accomplished by means of geometry, and in part achieved by mere scheduling of the constituent measure-ment operations of the range runthrough. Of course, receiver-gain control and/or transmitter-power control could also be utilized, if desired.
According to a fur~her concept of the invention, instead of proceeding through the range-steps in the order from the lower to the upper limit of the total range of interest, the runthrough proceeds in the opposite direction. Such downward runthroughs alternate with upward runthroughs, or are occasionally thrown in after every few up~ard runthroughs. As explained belowr the downward runthroughs more so than the upward runthroughs, but most especially in combination with upward runthroughs, avoid the generation of false ceiling values attributable to low-lying visibly transparent haze wrongly interpreted as low-lying clouds.
The novel fea-tures which are considered as charac-teristic for the invention are set forth in particular in the appended claims. The invention itself, however) both as to its construction and its method of operation, together with additional 1 objects and advantages thereof, will be best understood from the following description of specific embodiments when read in con-nection with the accompanying drawings.
FIG. 1 is a graph showing the reflected-pulse re-ception interval and the receiver gating interval, for the case of upward and downward runthroughs through the series of range-steps of a total ceilometric range to be explored;
FIG. 2 is a schematic diagram of the transmitter and receiver, illustrating a quadratically increasing beam pattern overlap; and FIG. 3 is a schematic block diagram illustrating the type of system configuration used to perform the inventive method. :~
One problem addressed by the present invention is preventing the misinterpretation of low-lying transparent haze as being non-transparent low lying clouds. This problem is addressed -by the present invention, by running through the range-step series not in the customary zero to upper-limit direction, but instead in the opposite direction. When, in conventional manner, the gating interval of the receiver is shifted for each new range-step, and the reflected laser pulses received during that range-step interval (if any) are totalized or integrated, the resultant signal can achieve a sufficient strength for signal-processing purposes, even if the peak power of each individual emitted laser pulse is re--latively low. However, to e~clude the effects of pulse reflections from low-lying but visibly transparent haze or smog, range-step runthroughs are performed in the downwards direction, e.g., one downward runthrough followecl by one upward runthrough performed in alternation~ or one downward runthrough thrown in after every few upward runthroughs.
If the ceilometric apparatus is designed for a 1 ceiling range of 3,000 m, corresponding to a travel time of 20 microseconds, then, in accordance with this concept of the invention the first range-step employed in the range-step runthrough is the 20-microsecond range-step, the schedule proceecling from there downwards towards zero, at an advancement rate such that the time required for the whole runthrough mee-ts interna-tional regulations, e.g., takes no longer than 15 secondsO In this way, first the highest-level detectable reflections are registered, then the next-highest, and so Eorth. When a predetermined number of r~-flecting layer heights have been detected, e.g., two, the down-wards runthrough can be terminated prior to completion; if high~
level reflecting layers are being detected, this can only happen if the haze or cloud layers located lower are so transparent that the laser radiation passlng thorugh them is scarcely attenuated.
Accordingly, the lower-lying haze or cloud layer heights would be sufficiently transparent to human vision and therefore not pro~ide an obstacle to vision for a pilot landing his aircraft.
If, for metereological purposes, it is also deslred to generate information concerning lower-lying thin haze or cloud layers, then, after performing the downward runthrough from 20 ; microseconds down to zero, one can perform an upward run-through starting from zero and going upward to the 20-microsecond range-step.
Indeed, this combination of downward and upward runthroughs in itself can yield further information, as explained with reference to FIG. 1. In FIG. 1, tx is the gating-interval location, on the time axis, for the uppermost range~step, e~g., the 20-microsecond range step. The gating interval T, of duration tT is being shifted downwards in direction A, i.e., from the upper range-limit down towards 2ero. The reflected laser pulse S of 1 duration tSl is being applied to the receiver, but the range-step ~or wh ch the gating interval T has been se~ at this point is higher than the ac-tual height of the reflecting cloud, and the reflected pulse S applied to the receiver is not being applied during the gating interval T. As one shifts more and more down-wards through the range-steps, the gating interval T wilL shift in direction A towards the interval of received signal S. As this point first begins to be reached, the edge KT of the gat:ing interval T begins to overlap the edge KS of the reflected laser pulse. If this pulse S is being reflected off a high cloud, the receiver, when this overlap begins, is not able to respond to the very low signal strength, and not until the two intervals tsl and tT
completely coincide are enough photons being received at the gated receiver for the latter to begin to respond and register.
In contrast, with low-lying clouds, the reflected energy is of course much greater, and therefore already at the beginning of this overlap the strength of the signal applied to the receiver will be quite high, and the response to and regis `
tration of this signal will not wait for the two intervals to ~r completely coincide. This can lead to the generation of mis-information concerning the exact height of low-lying clouds; and it is low-lying clouds whose exact height is most critical to know. For example, in the case of a zero-ceiling ground cloud, and if the transmitted laser pulse has a penetration depth of e.g., 50 meters into such cloud, the registered and indicated height of this zero-ceiling cloud will be 50 meters in addition to the distance corresponding to the duration tSl of pulse S. For example, if tSl is 100 nanoseconds, corresponding to a trans-mission~reflection travel time of 15 meters for the laser pulse, the indicated cloud height would be 50 ~ 15 = 65 meters, whereas 3r7~i~

l what is in truth involved is a z~ro-ceiling cloud having a visi bility depth of 50 meters, a typical value for rain clouds. The system should have indicated that the cloud height is zero.
If now, one first performs a runthrough in direction A and then another runthrough in direction B, i.e., first a down-ward runthrough and then an upward runthrough, exactly the opposite will occur. I.e., for a low-lying cloud, as soon as the B-direction-shifted gating interval T begins to overlap the re~lected pulse S, a high signal strength is encountered, and the indicated cloud height is lower than actual. By combining both runthroughs, the recorder registers as a result of the A-direction runthrough the vertical visibility depth through the cloud; this information can be superimposed upon that generated by the B-direction runthrough, so that the latter runthrough properly yields an indication of a zero-height ground cloud. Of course, because the A-direction runthrough is still being used, the ready detectability of high-level clouds is fully preserved. Accoringly, this combination of downwards and upwards runthroughs can be made to meet the re-quirements of both metereological and flight-safe~y-system appli-cations.
International safety regulations pertaining to eye- -safety standards for all such laser measuring systems all demand that a person, who for whatever reason looks directly into the oncoming laser beam~ suffer no eye damage. These standards are ; extremely strict and conservative, and have forced the development of such ~easuring systems into the use of very low laser-pulse power levels, but combined with pulse-repetition rates higher than would otherwise be employed, e.g., lO Watts power, 5,000 pulses per second, and 200-nanosecond pulse durations. With pulses of such low power levels, integration of multiple pulses received (if y~

l any~ during each individaul range-step must be resorted -to, e.g. r totalizing for example about lO0,000 pulses during the gating intervals of the receiver sys-tem.
Necessarily, this tends ~o limit ceiling range measurements to abou-t 1000 meters, furthermore with the condition that the air below the clouds at such levels be quite clear.
Attempts to merely increase the laser power are considered im-possible, because there would be no way then to construct the system to meet eye-safety standards. However, the formulation of the international standards in question is generally such as to have led to two ways to deal with this problem.
1. Decrease the speci~ic power of the emitted laser radiation, .i.e., the power per unit surface area of bea~ cross-section. This can be done by employing at the transmitter com-paratively large optics, e.g., lenses or mirrors having diameters on the order of 40 cm. Preferably, the receiver system is then pro~ided with correspondingly large optics, in order to receive as : much of the thusly thinned reflected radiation as possible.
2~ ~ecrease the repetition rate of the emitted laser pulses, in order to be able to use higher per-pulse power without exceeding average-power limits. However, this inherently increases the time required for the range-step runthrough. For example, to run through a practical range of ceiling values might then require ;~ a time interval on the order of minutes, ~Ihereas international flight-safety-system standards require that four fresh measurements be generated per minute, thus leaving only 15 seconds for each measurement.
The present invention contemplates a scheduling technique, such that the number of reflected received laser pulses (if any) integrated per range-step varies in dependence upon which 3~

1 ran~e-step is involved. The purpose of the multiple integrations per range-step is to generate a signal of suEficient strength Eor reliable signal-processing. Inasmuch as -the pulses reflected from high-level clouds are much weaker than those reflected from lower-level clouds, fewer integrations are needed for range-steps in the latter case than in -the former case. This can be done in several ways.
First, the invention contemplates varying the re-petition rate of the emitted laser pulses as a ~unction of range-step. Thusl for example, when working at the 3000-meter range-step, the laser-pulse repetition rate could be 10 kHz, 2 kH~ at the 1500-meter range-step, for still lower range-steps only a few hundred Hz, and then for range-steps at e.~., 300-500 meters only one or two pulses. Preferably, the functional dependence o~ the repetition rate on range-step is progressive and appro~imately quadraticO
The reason for the quadratic relationship will be understood by persons skilled in the art. The power loss o~ the transmitted and re~lected pulse radiation theoretically corre-sponds to the square of distance, i.e., range-step value. Actual-ly, the repetition rate of the laser pulses can be decreased, when going in the direction from higher range-steps to lower range-steps, faster than quadratically. This i5 because, besides the quadratic dissipation of power, there is also dissipation of power by simple absorption in the air. Accordingly, relative to the repetition rate used for the highest range step, that for the exactly middle range-step can be lower than 25~, etc. Thus, the repetition-rate of the emitted laser pulses is progressively decreased during a downward runthrough or, equivalently, pro-gressively increased during an upward runthrough. The rate at $~

1 which the gating interval of khe receiver is shifted, however, remains constant.
When, in this way, only the pulse repe-tition rate is varied, but the gating-interval shift ra-te kept cons-tant, the number of pulses emitted fo.r the entire range runthrough .is reduced and accordingly, for a given per-pulse energy, the average power per runthrouyh is lowered. However, little is done to reduce the time required Eor the runthrough.
Therefore, a second contemplated scheduling technique, 10 which likewise varies in dependence upon range-step the number of received pulses (i:f any) integrated per range step, involves -the progressive variation in the shift rate of the gating interval.
I.e., the time spent working at each single range-step progressively changes, as one proceeds through the succession of range-steps. r In the extreme form of this alternative technique, the repetition rate of the emitted laser pulses is constant -throughout. However, for each range step, it is quite accurately known in advance how many reflected and received pulses would have to be integrated for that range-step, i.e., assuming that reflected pulses are received during the gating interval of that range-step, in order to build up a signal strong enough for signal-processing and, more general-ly, a signal wich is normalized.
Accordingly, with the repetition rate of the emitted laser pulses constant, the time allotted to each range-step is such that, if pulses are received during the gating interval of that range-step, there is exactly time enough to receive a number of such successive reflected pulses sufficient to build up a signal of sufficient strength. I.e., the number o~ in-tegrations performed during each range step, or a-ttempted in the event no signals are received, progressively increases as one proceeds from 1 lower to higher range-steps, preferably in the approximately quadratic manner jus-t referred to with respect -to the repetition-rate-variation alternative.
In either case (progressively varying repetition rate, but ga-ting-interval shift rate constant; constant repetition rate, but ga-ting-interval shift ra-te progresssively vary:ing), the number of pulse integrations to be performed per range-step pro-gxessively varies, preferably at a somewhat faster than quadratic rate when proceeding in the downwards direction through the suc-cessive range-steps.
Of course, combinations of these two techniques would also be possible. Clearly, however, for -the sake of simplicity, it is preferable to use only one or the other, where the number of range-steps and the resolution per range-step permits this.
In general, with ceilometric systems o the type in ;
question, it is typically considered necessary tha-t the gain of the receiver be varled as a quadratic function of range. This requirement relates both to operativeness per se and also, some-what more generally, to signal normalization throughout the range explored, both of these having been discussed aboveO The imple- ~;
mentation of a quadratic gain for the receiver per se increases the number of electronic components which must be involved for signal-reception; this in turn leads to decreases in the mean time between failures (MTBF)~ However, a high MTFB is of course a stringent requirement for flight safety systems.
The present invention contemplates achievement of the equivalent oE such a quadratic gain1 but without necessarily using actual quadratic-gain amplifier cQmponents in the receiver. The explanation already made of how the number of integrations per-formed per range-step is to be varied as a quadratic function of 7~

l range represents a partial solution to the problem, but not yet a complete solution. To appreciate this~ the following can be considered.
With the first alternative described above, the gating-interval shift rate is kept constant, but the repetition rate of the emitted laser pulses is progressively varied, ap-proximately quadratically. This can create problems when trying to span the transition between the upper and lower parts of the total range of interest, the total range of interest being e.g., 5000 ft down to 20 feet. Depending upon the size of the range and ~ ;
the value allotted to each range-step, the repetition rate will go 'r down as one runs down through the range-steps. For example, if at the upper limit of the range the repetition frequency is to be 104 EIz, then at the lower end of the range of interest the repetition frequency of the emitted pulses must have gradually decreased to ~-frequencies on the order of 1 H~. Of course, this is an extremely low repetition frequency, and prohibitively prolongs the time that would be required to run through the whole series of range-steps.
A similar kind of problem can arise with the second alternative discussed above, i.e., wherein the repetition rate of emitted pulses is constant, but the gating-interval shift rate is progressively varied. At the lower end of the succession of range-steps, there is virtually no need for multiple integration of received reflected pulses, because their energy is still so high. If therefore, there is allotted only a single in-tegration for the lowest range-step, then the number of integrations per formed as one progresses upwards through the successive range~
steps should go up quadratically as a function of height. How-ever, going, for example, from the first (lowest) range~step to the second range-step, the quadratic increase in the num~er of 1 integrations performed is not readily implementedO If one inte-gration is performed for the first range-step, then to perform for the second range-step again only one integration would~ stric-tly speakin~, not be enough, whereas two integrations would be too many, i.e., at least in principle.
Clearly, this results from the rate of change of a quadratic function; the rate of change progressi~ely increases from initially low values to higher and higher values. To try -to allot to each range-step the proper number of integrations is therefore problematic in the lower part of the total range.
Obviously, to attempt "partial" integrations would be extremely complex and would destroy the integer-based character of the operating cycles. If one tried to allot to the lowest range-step a higher number of integrations, e.g., 10, in order that there would be more freedom in approximating to a guadratic increase when allotting the number of integrations for the next-lowest range-step, then~ clearly, by the time the highest range-step had been reached, the number of integrations performed would be ex-tremely great.
This problem does not exist at the upper part of the total range of interest. There, the number of integrations performed per range-step is high, and therefore, when going from one of these range-steps to the next, a fractional increase in the number of integrations performed, such as to approximate to a quadratic increase, can readily be implemented.
The desired quadratic increase, it should be re-called, has two aspects. One is operativeness per se, in the sense that the receiver must generate processable signals as one goes into the higher parts of the ceiling range. The related, but distinct aspect is signal normalization. Although the quadratic 1 increase, and the problem of i~plementing it with the two al-ternative techniques described, is particularly troublesome when ai~ing for ideal signal normalization, it is encountered to a lesser degree, but still to a degree which must be dealt withl at the level of operativeness per se.
Thus, tlle problem of bridging across the transition between the lower and upper parts of the total ceiling range of interest is a very real one and is, in a sense, one importan~
reason why ceilometric systems meeting the needs of both fligh-t-safety and ~etereological applications are problematic.
The present invention hridges this transition usingthe technique already described (progressive variation o~ the number of integrations per range-step) plus a simple mechanical technique for implementing the desired quadratic relationship in the lower part of the total range of interest. In particular, the slightly conical beam or radiation patterns of -the transmitter and receiver are permitted or caused to intersect and thereby overlap, with re&pect to their cross-sectional areas, to an exten-t which increases with increasing height. The beam diagrams of the trans-mitter and receiver are, to begin with, slightly divergent, i.e.,conical. To facilitate the establishment of this increasing overlap, the light-emitting and receiving components at the trans-mitter and receiver can be positioned slightly decentered relative to their optical axes. This is shown in FIG. 2. The basic con-struction of the transmitter unit T and of the receiver unit R are conventional. The transmitter ~ comprises a laser light source LS, and the receiver R a light detector LD.
These components each cooperate with a respective ~yperbolic mirror HM and a respective parabolic mirror PM.
However, preferably both these components are posi-tioned slightly 3~

1 Off the optical axes of their respective optical systems, to facilitate the beam overlap. The transmitter and receiver beam paths begin to intersect at an elevation of e.g. 10 meters. rrhe overlap of the two beams progressively increases, until at a height o~ e.g., 300-500 meters maximum overlap is reached. The cross-sectional area of the overlap of two thusly intersec-ting cones is itself proportional to the square of height. In this way, by simple mechanical means involving geometry and optics, a quadratically increasing sensitivity at the receiver is achieved.
Beyond the height of maximum overlap, the continuance of the desired generally quadratic sensitivity relies upon the already described technique of progressively varying the number of pulse integrations performed per range-step. The progressive increase in the number of integrations can begin for range-steps above the region of maximum beam overlap, or can begin already at lower range-steps.
FIG. 3 is a simplified schematic block diagram of the electronics of such a ceilometric instrument. The light-emitting component of the transmitter is a GaAS or GaAlAs diode array of very rectangular construction and a maximum power of e.g. 400 WO
The diode array is driven, however, with only 200 W peak power and with a pulse duration of 70-90 nanoseconds. The si~e of such an array is approximately 3 x 3mm. The array must always be somewhat larger than the circle of confusion of its cooperating transmitter optics~ By driving the diode array at below maximum permissible power, and using pulse lengths shorter than the maximum permissi-ble value of 200 nanoseconds, it becomes possible to avoid the so-called Periot-Fabry erosion which would otherwise occur at the edges of the individual laser diodes of the array. Also, care should be taken that -the temperature of the diode array be kept at -- ~0 --~3~

1 least at a certain minimum, e.g., 20C, in order -tha-t the electrl-cal pulses applied to the array not destroy the diodes, which can happen at low tempera-tures due to the high efficiency of the diodes at such temperatures. Thus, in the preferred embodiment, cooling of the laser array is not employed, and instead heatin~ is reso~ted to, e.g., electronically regulated temperature control.
The electronic generation of the pulses which drive the laser diode array can be performed in conventional manner using a thyristor or transistor control circuit, pre~erably utilizing a delay-line network to make the leading and trailing flanks of these pulses as sharp, and therefore the pulses -themselves as rectangular, as possible. However, because such -thyris-tor pulse-generating circuits emi-t extremely large amounts of electronic interference, the entire circuit i9 preferably provided in a housing made of soft magnetic material serving to completely shield in the high emitted magnetic surges. However, such a housing, which may be generally rec-tangular, should include a copper or aluminum plate of sufficient thickness, on which are mounted the heat-generating electronic components, so that the generated heat can be conductively transmitted to the ambient air.
The entire housing should be completely shielde~ in to prevent the emission of high-frequency electrical fields, using for example so-called contact metal (brushlike, sharp-edged copper braiding) at the junctions of the housing component. This serves to prevent any measurable electrical inter~erence fields generated by the laser transmitter from finding their way into the e~tremely sensitive receiver unit.
This expedient is in particular necessary when de-tecting very low-lying clouds, because otherwise not yet decayed interference pulses could become superimposed upon the returning 37~

1 reflected laser pulses, these returning af-ter a few nanoseconds.
This would lead to persisting indications oF ground-level clouds.
The radiation detector used in -the receiver is preferably either a silicon-avalanche diode or a simple Schot-tky-barrier diode, again it being necessary that the effective surface of the radiation-receiving component be greater than the circle of confusion of its cooperating optics. After preampliEication of the received signal, e.g., using a low-noise MOSFET, the pre-amplified signal is applied to a so-called sampling bridge, which establishes the shift for the receiver's gating interval, as already described above. Only if the preamplified signal exists during the ga-ting interval established for the range-s-tep involvedt does the siynal reach the next electronic signal~pxocessing stages.
If it does, it is first amplified once more, and then applied to an integrator. It will be understood that an integrator can detect signals even when such signals are buried within noise generated by the signal-processing circuitry itself, io e., self-noise.
If the amplitude of the signal accumulated by the integrator clearly enough exceeds the noise level, then a time value (indicating what range-step is involved) is stored in a memory, indicating that a cloud reflex has been encountered at this range-step. The range-step at which a reflex has been thusly detected is also indicated on a display, or the like, for a viewable reading of the ceiling value, this indication persisting until, during the next range step runthrough a cloud reflex, perhaps at a different range-skep, is again encountered~ whereupon the new value is now indicated, or the old value continues to be displayed.
The integration schedule for the successive 7~

1 range~steps has already been described at len~th above. Prefer-ably, the schedule is stored in and implemented by a microprocessor, which then can readily be employed to perform auxiliary operations which may be useful in the con-text of such instruments. Chief among these auxiliary operations is calcula-tion of the percen-tage of time, e.g., duriny an ongoing time interval of 30 minutes duration, for which clouds were present above the ceilometer at ranges of 100-200 meters, 200-~00 meters~ ~00-800 meters~ etc.
This data is stored in an ongoing manner and can be periodically read out, to characteri~e the cloud-cover situation at a particular area; this informa-tion is of-ten desired by amateuer pilots and for metereological applications.
Advantageously, such microprocessor can -then also be used to automatically adjust the detector threshold level (this is not the same as gain), such that as the noise in the signals downstream of the sampling bridge ~gating interval establishing stage) increases or decreases, the threshold level for signals is correspondingly adjusted, so that when signals are registered, after they have been subjected to (mul-tiple) integration they will have a value just somewhat above the noise value, irrespective of whether the noise amplitude is large or small. The chief source of such noise is the high D.C. current component resulting from exposure to ambient daylighi, which becomes superimposed upon -the signal pulses in the photodiode.
Although the general type of constructions which are preferably used for the transmitter-receiver arrangement are known, reference may be had to FIG. 3 and the accompanying de~
scription in German Federal Republic published patent application DT-AS 2,150,969 for background information.
In accordance with a further concept of the present 3~

1 invention, use is made of a self-check feature, serviny to -test the operativeness of the instrument once per runthrough through the successive range-steps. Preferably, this is performed as follows: The end of a light-conductive element is located in -the path of the emitted laser pulses, for reception of a very small fraction of the emitted light. The other end of the light-con-ductive element emits such light onto a photodiode in the receiver, this other end being located within the housing of the receiver.
The photodiode ~uxnishes a corresponding signal, possibly through the intermediate of an ampliEier~ to a time-delay stage having a time-delay interval of e.g. 20.1 microseconds, i.e., corresponding to a range-step one step above the highest range-step in the succession of range-steps. The time-delayed signal is then applied to an infrared LED whose emission wavelength range is the same as that of the laser-diode array used at the transmitter. Accordingly, once per runthrough, just above the highest range-step ~correspond--ing to a travel time of e.g. 20 microseconds), a signal is constant-ly furnished corresponding to a still higher range-step Ihere 20.1 microseconds). If, for example, a recorder is connected to the system, a corresponding small stroke will be registered at the edge of the chart. The recorder may have a measuring range of 0 to 7500 feet, and the extra range-step for self-check purposes, being a 50~foot range step, is readily detected on the recorded chart. If the self-check marking on the recorder chart is absent, then this indicates that during the associated range-step run-through the instrument was either switched off or mal functioning.
As is well known, laser diodes have the tendency, to emit progressively higher wavelengths as their temperatures pro gressively rise. This can make it necessary to give a grea-ter than otherwise necessary bandwidth to the narrowband optical ~3~

1 filter covering the receiver's photodiode, this narrowband filter being provided to isolate the receiver's photodiode from the effect of ambient daylight. Therefore, if one wishes to keep the bandwith of the narrow band optical filter as small as possible, it becomes necessary, at considerable expense, -to employ electronic-controlled Peltier elements to cool the laser~diode array, and also to heat it when its temperature falls too low, so as -to stabilize the temperature of the array. In order to avoid this problem with less expense, the present invention prefers to employ an array comprised of multi hetero-structure GaAs or GaAlAs material.
This material exhib.its only a very low temperature-dependence in the wavelength of its emitted radiation, making it possible to dispense with cooling.
If such material is employed, it then suffices, even for example upon exposure o F the instrument to strong summer sunlight, to employ a high-thermal-conductivity rod heat-conductive-ly connecting the housing of the laser light source to the metallic base of the instrument, which latter because of its proximity to the ground is sufficiently cool.
The scheduling techniques contemplated by the present invention have been elucidated at length above. Returning to the question of eye-sa~ety standards, it will be noted that progressive dec.rease in the repetition rate of the emitted laser pulses serves directly -to cut down the average emitted power of a range-step runthrough, but does not per se lower the time required for the runthrough. In contrast, if the repetition rate of the emitted pulses is kept constant, and the number of integrations performed per range-step is progressively varied by varying the gating-interval shift rate as a function of range-step, then this speeds up the runthrough very considerably, but does nothing per se to reduce a~erage em.itted power. Accordingly, when use is made of 3~

1 the second technique, it is preferred to introduce pauses of operation~ at the ends of runthroughs and/or scattered into -the runthroughs themselves, in order to thusly reduce the average emitted power. These pauses in operation can then be also utilized for performance of signal-processing operations, such as transfer of accumulated data from a buffer storage to the memory of a microprocessor and preparation of the buffer storage for the receipt o:~ new data, etc.
It will be understood that each of the elements described above, or two or more together, may also find a useful application in other types of constructions and procedures differing from the types described above.
While the invention has been illustrated and described as embodied in a particular combinations of schedulin~ procedures with device geometry, to yield generally quadratic receiver operation, it is not intended to be limited to the details shown, since various modifications and structural changes may be made without departing in any way from -the spirit of the present in-vention 4

Claims

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. An improved ceilometric method of the type in which a transmitter is used to transmit pulsed laser light upwards for subsequent downwards reflection off of cloud bodies, the thusly reflected laser pulses being incident upon a receiver, the travel time of each such transmitted and received reflected pulse being dependent upon the height of the reflecting cloud body, the total range of ceilometric values to be explored being subdivided into a succession of range-steps each corresponding to a different laser-pulse travel time, the exploration of each individual range-step being performed by gating the receiver using a gating interval whose location in time is shifted relative to the time of emission of an exploring laser pulse by an amount corresponding to the laser-pulse travel time associated with the respective range-step, a runthrough through the successive range-steps being performed by progressively shifting the receiver's gating interval with respect to emission time, the reflected pulses received by the receiver during the gating interval associated with the individual range-step currently being explored being applied to a signal integrator to build-up a signal of magnitude sufficient for signal processing, the improvement comprising (a) dividing said total range of ceilometric values to be explored into a lower part and an upper part, and for the range-steps in said lower part establishing a transmitter to receiver transmission factor which progressively increases in the direction of higher and higher range-steps, this being done by orienting the conical upward beam patterns of the transmitter and receiver such that proceeding in the upward direction they inter-sect and increasingly overlap, the area of beam pattern overlap increasing approximately quadratically in the upwards direction, whereby to take into account by means of geometry the range-dependent power attenuation suffered by laser pulses in travelling from the transmitter upwards to a cloud body and then down to the receiver; and (b) at least for the range-steps in said upper part of said range, establishing an integration schedule fixing, for each single range-step, the number of transmitter-emitted pulses to be integrated for that range-step before going on to another range-step, the number of integrations performed per range-step progressively increasing in the direction of higher and higher range-steps, whereby to take into account by means of progres-sively increasing numbers of repeated pulse integrations the range-dependent power attenuation suffered by laser pulses in travelling from the transmitter upwards to a cloud body and then down to the receiver.

2. In a method as defined in claim 1, the number of repeated pulse integrations performed per single range-step progressively increasing from values on the order of one to values on the order of 100.

3. In a method as defined in claim 1, the progressive increase in the number of pulse integrations performed per range-step comprising establishing a repetition-rate schedule for the transmitter-emitted laser pulses such that the repetition-rate of emitted pulses is a progressively increasing function of increasing range-step value.

4. In a method as defined in claim 3, the rate at which the gating interval of the receiver is shifted when proceeding from one range-step to the next during a runthrough of the range-steps being constant.

5. In a method as defined in claim 1, the progressive increase in the number of pulse integrations performed per range-step comprising progressively varying as a function of range-step the rate at which the gating interval of the receiver is shifted when proceeding from one range-step to the next during a run-through of the range-steps.

6. In a method as defined in claim 5, the repetition rate of the pulses emitted from the receiver being constant when proceeding from one range-step to the next during a runthrough of the range-steps.

7. In a method as defined in claim 1, the progressive increase in the number of integrations performed per range-step being an approximately quadratic increase proceeding in the di-rection of higher and higher range-steps.

8. In a method as defined in claim 7, the runthrough through the whole succession of range-steps being performed in the direction from the highest-value to the lowest-value range-step.

9. In a method as defined in claim 1, the runthrough through the whole succession of range-steps being performed in the direction from the highest-value to the lowest-value range-step.

10. In a method as defined in claim 1, the overlap of the conical beam patterns of the transmitter and receiver progressively increasing in the upwards direction to a height of maximum overlap, the height of maximum overlap being on the order of 1000 feet.

11. In a method as defined in claim 10, the height of maximum overlap being at least 500 feet.

12. In a method as defined in claim 5, the time required for a complete runthrough through the succession of range-steps being a fraction of a minute, further comprising the step of introducing, prior to the next such runthrough, a dead interval during which no laser pulses are emitted by the trans-mitter, in order to reduce the average power of the emitted laser pulses relative to the average power which would obtain if the next runthrough were immediately commenced.

13. In a method as defined in claim 1, the steps at (a) and (b) being the exclusive means utilized for progressively varying the transmitter-to-receiver transmission factor, the per-pulse output power of the transmitter and the amplification of the receiver being maintained constant through the whole runthrough.

14. In a method as defined in claim 1, furthermore comprising checking for operativeness once per runthrough, this comprising applying to the receiver once per runthrough a pulse of laser light whose wavelength, strength and timing simulates that of an actually transmitted laser pulse reflected off a cloud whose height approximately corresponds to the highest range-step.

15. In a method as defined in claim l, using for the laser-pulse emitting component of the transmitter a component comprising multi-hetero-structure GaAs material.

16. In a method as defined in claim 1, using for the laser-pulse emitting component of the transmiter a rectangular array of laser diodes.

17. In a method as defined in claim 16, energizing the laser-pulse emitting component of the transmitter using electronic circuitry confined in a housing of soft magnetic material and dissipating heat generated by such electronic circuitry to the outside of such housing using a heat conductor of thermally conductive metal.

18. In a method as defined in claim 1, using a microprocessor to store and implement the integration schedule.

19. In a method as defined in claim 18, furthermore using such microprocessor to automatically adjust the response threshold of the components in the receiver to which the inte-grated pulses are applied in dependence upon noise level.

20. In a method as defined in claim 18, furthermore using such microprocessor to calculate and retrievably store the percentage of time per extended unit period of time during which clouds are above the transmitter and receiver, for a plurality of different cloud height ranges.

21. In a method as defined in claim 1, establishing the beam pattern overlap using a transmitter comprising a laser-light emitting component and transmitting optics and using a receiver comprising receiving optics and a laserlight receiving component, the emitting component and the receiving component each being maintained decentered relative to the optical axis of the associated optics.

22. An improved ceilometer of the type comprising a transmitter and a receiver, the transmitter emitting pulsed laser light upwards for reflection off of overhead cloud bodies and subsequent downward travel onto the receiver, the travel time of each such transmitted and received reflected pulse being dependent upon the height of the reflecting cloud body, the total range of ceilometric values to be explored being subdivided into a succession of range-steps each corresponding to a different laser-pulse travel time, the receiver including gating means establishing which range-step is to be explored by establishing for the receiver a gating interval whose location in -time is shifted relative to the time of emission of an exploring laser pulse by an amount corresponding to the laser-pulse travel time associated with the respective range-step, the ceilometer including means performing a runthrough through the successive range-steps by progressively shifting the gating interval with respect to emission time, the receiver including pulse-integrating means connected to receive and integrate pulses received during the gating intervals to thereby build up signals of magnitude sufficient for signal processing, the improvement wherein:
(a) the transmitter and receiver are oriented such that their conical upward beam patterns intersect and increasingly overlap proceeding in the upward direction, the increasing overlap being progressive and an approximately quadratic function of height, the height of maximum beam pattern overlap dividing the total range to be explored by the ceilometer into a lower and an upper part, this quadratic increase in the amount of beam pattern overlap serving to compensate for the range-dependent power attenuation suffered by laser pulses in travelling from the trans-mitter upwards to a cloud body and then down to the receiver; and (b) scheduling means operative for automatically implementing an integration schedule which fixes, for each single range-step, the number of pulse integrations performed before going onto another range-step, the number of integrations per-formed per range-step progressively increasing in the direction of higher and higher range-steps.