WO2017134398A1

WO2017134398A1 - Device for predicting rain/snow line altitude

Info

Publication number: WO2017134398A1
Application number: PCT/FR2017/050246
Authority: WO
Inventors: Guy Badoche Jacquet
Original assignee: So.No.Par
Priority date: 2016-02-03
Filing date: 2017-02-03
Publication date: 2017-08-10
Also published as: FR3047375A1

Abstract

The invention relates to a rain/snow line altitude prediction device (1) which includes: - a memory (3) for receiving meteorological data sets, each meteorological data set combining a geographical portion (i), a time (h), a temperature (T), a precipitation indication (P), at least one altitude (K, H, IK, IH), and an air mass type (M) indicating if an air mass combined with the geographical portion (i) for a time (h) is considered to be entering or leaving a geographical area corresponding to the combination of the geographical portions (i), each meteorological data set being initialized with an air mass type (M) indicating a leaving air mass, and the times (h) being selected such that the meteorological data sets can be grouped by time (h) for a given geographical portion (i), - a sorter (5) arranged so as to search through a meteorological data set group including a current time (h) and, for a given set from said group, change the air mass type (M) to indicate an entering air mass when the difference between the temperature (T) of the given set and the temperature (T) of a set combining the geographical portion (i) of the given set and a time (h-2; h-1) prior to the current time (h) is greater than a threshold value (T1; T2), and - a calculator (7) arranged so as to calculate, at a given time (h), a predicted rain/snow line altitude (IK, IH) for each air mass type (M), each air mass being from a mean isothermal altitude (K, H) estimated for the relevant air mass type (M) for both times immediately prior to the given time (h) for geographical portions in which precipitation is detected at the given time (h). Said isothermal altitude (K, H), estimated for the given time (h) and a given air mass type (M), is calculated on the basis of the mean, for the meteorological data sets including the given time (h) and air mass type (M), of the sum of the altitude (z) of the relevant set and an altitude calculated on the basis of the temperature (T) of the relevant set and a vertical temperature gradient (DTH).

Description

Altitude prediction device for the rain-snow limit

The invention relates to the field of weather forecasting, in particular the automatic prediction of snow-to-snow limit altitudes and snow phenomena at low altitude.

It is known to detect the presence of hydrometeors in the atmosphere, that is to say water in various phases (liquid, solid - crystallized in various forms - or mixed). Dual polarization radars are the most common tools for making such detections. Nevertheless, this detection is done in the radar beam at a high altitude relative to the ground and at a given moment, and therefore does not prejudge the presence on the ground of solid or mixed hydrometeors. In addition, this detection can not be used as a short-term forecast. Even associated with an algorithm for measuring the displacement of hydrometeors, it does not predict the phase of hydrometeors that will fall in places of different altitude. In addition, other variables such as temperature instead of forecast may change as the hydrometeor moves.

It is generally accepted that knowing the altitude of the lower limit of the melting layer (ie the thickness of this layer below the altitude of the 0 ° C isotherm) makes it possible to estimate the nature of hydrometeors on the ground: when this melting layer is close to the ground, a snow phenomenon occurs because hydrometeors do not have time to melt before reaching the ground. Conversely, a melting layer located at a high altitude from the ground indicates a rainy phenomenon.

In winter and at altitude, for example above 1,000 meters, the soil is almost permanently above the melting layer. Therefore, all precipitation takes the form of snow. Studying the altitude of the melting layer is of little interest in this context. Predicting the presence of hydrometeors is enough to predict the snowfall. In the lowlands, for example below 1,000 meters of altitude, the forecast of the vertical temperature profile should theoretically make it possible to predict the snow phenomena in time and space. The forecast of the evolution of the vertical temperature profile in the short term, for example in the next 12 hours, is very difficult, particularly because of rapid changes in the altitude of the melting layer in the snow phenomena "by isothermal Rainfall occurs primarily as rain with low positive air temperatures near the ground. At altitude, the melting of the falling snow is associated with a latent heat release which induces a cooling of the air. As a result, the rain / snowline lowers, melting occurs more and more close to the ground and eventually reaches it. Precipitation then occurs in plain in the form of snow.

In order to provide predictable reliability, the initial input conditions of existing forecast models require frequent updates to account for rapid changes in the elevation of the melting layer. However, the models used, in particular the "AROME" model used by Météo-France, have computation times of about 3 hours, plus the interpretation time of raw data by meteorologists. It is generally considered that there is a window of at least 4 hours during which the presence of snow is unpredictable. Only snow sensors on the ground are then effective to detect (necessarily a posteriori) the snowfall.

According to a second approach, particularly implemented in Switzerland, the ground temperature, updated every 10 minutes at a given point, and the immediate forecast by extrapolation of the radar images, updated every 5 minutes, are coupled. A probability of precipitation (fall of hydrometeors) by geographical area is drawn from the extrapolation of the radar images. Simultaneously, an autoregressive model using a standard vertical temperature gradient of 0.65 ° C / hr and an estimate of relative humidity based on previous measurements allows a solid form probability of water to be calculated at the point of measurement at from the temperature on the ground. The product of the precipitation probability and the solid form probability of the hydro meteors gives a probability of snowfall. The results are interpolated to any desired geographic point from the geographical points for which measurements are available. This second approach has an unsatisfactory reliability. No device or method therefore allows a reliable forecast of snow in the short term and at low altitude. This limits the effectiveness of preventive and curative measures, particularly in the field of road and airport safety.

The invention improves the situation.

The Applicant proposes a device for forecasting the altitude of the rain-snow limit including a memory for receiving sets of meteorological data, each set of meteorological data associating a geographical portion (i), a moment, a temperature, a precipitation indication. , at least one altitude, and a type of air mass indicating whether an air mass associated with the geographical portion for a moment is considered as entering or leaving a geographical area corresponding to the union of geographical portions, each a meteorological data set being initialized with an air mass type indicating an outgoing air mass, and the instants being chosen so that the meteorological data sets can be grouped by time for a given geographical portion, a sorter arranged to browse a group of meteorological datasets including a current moment, and, for an e Given this group, change the type of air mass to indicate an incoming air mass when the difference between the temperature of the given set and the temperature of a set associating the geographical portion of the given set and a time prior to the current time is greater than a threshold value, and a calculator arranged to calculate, at a given instant, an expected altitude of the rain-snow limit for each type of mass d , each from an estimated average isothermal altitude for the respective air mass type for the two instants immediately prior to the given instant for geographical portions in which precipitation is detected at the given moment , which estimated isothermal altitude for the given moment and a given type of air mass is calculated from the mean, for the meteorological data sets including the given instant and type of air mass, the sum of the altitude of the respective set and an altitude calculated from the temperature of the respective set and a vertical temperature gradient.

Such a device makes it possible to establish a forecast of snow phenomena in the lowland, at low altitude and in the short term with satisfactory reliability. Such forecasts allow roadside services to accurately target snow-covered geographic areas within the next 4 hours. This allows interventions upstream of the first flakes, or at least very quickly after their fall.

These forecasts more generally allow a better organization of the interventions according to the planned events. Such a prediction device can feed a centralized system for steering intervention teams, advantageously assisted by a geolocation system. Such a device makes it possible to fill the absence of intermediate information between the long-term and low-precision forecasts, and the snow detectors, by definition reliable but inherently of an inherently late nature. Short-term forecasting can be used in conjunction with long-term forecasting to adjust the level of alertness of response services.

The proposed device may have one or more of the following features, alone or in combination:

the altitude calculated from the temperature of the respective assembly and from a vertical temperature gradient corresponds to the product of the vertical temperature gradient by the difference between the temperature of the respective assembly and the isothermal temperature, altitude corresponds to that of the rain-snow limit,

the sorter is arranged to determine whether the type of air mass must be modified by comparing the temperature of the given set and, on the other hand, the temperature of a set associating the geographical portion of the given set with the instant immediately before the current moment, the sorter is arranged to determine whether the type of air mass must be modified by comparing the temperature of the given set and, on the other hand, the temperature of a set associating the geographical portion of the given set with the second instant immediately before the current moment,

the sorter is further arranged to modify the type of air mass to indicate an outgoing air mass of the meteorological data sets comprising a given instant, when all the data sets comprising the instant immediately preceding said given instant present a type of air mass indicating an incoming air mass,

the estimated isothermal altitude for each type of air mass results from a linearization of the evolution of the average isothermal altitude values at the previous instants,

the computer is furthermore arranged to reiterate the calculations, for instants after the current instant, after the calculations for the current moment,

the calculator is further arranged to correct the estimated isothermal altitude for a given moment for the type of incoming air mass, when a temperature lower than the isothermal temperature whose altitude is sought is detected at the given moment,

the device further comprises a comparator arranged to compare the estimated isothermal altitude for a moment and a geographical portion at a ground elevation of a location of said geographical portion, and to deduce a prediction of a snowfall at said location; at the moment when said altitude is higher.

The invention further relates to a method for predicting the altitude of the rain-snow boundary comprising:

at. receive sets of meteorological data, each set of meteorological data associating a geographical portion, a time, a temperature, an indication of presence of precipitation, at least one altitude, and a type of air mass indicating whether an air mass associated with the geographical portion for a moment is considered as entering or leaving a geographical zone corresponding to the union of the geographical portions, the instants being chosen in succession, so that the meteorological datasets can be grouped by moment for a portion geographical location, b. initialize each set of meteorological data with a type of air mass indicating an outgoing air mass,

vs. traversing a group of meteorological data sets including a current instant, and, for a given set of this group, modifying the type of air mass to indicate an incoming air mass when the difference between the temperature on the one hand of the given set and secondly the temperature of a set associating the geographical portion of the given set and a time prior to the current instant is greater than a threshold value,

d. calculating an estimated isothermal altitude for a given moment and type of air mass from the mean, for the meteorological data sets including the given moment and type of air mass, the sum of the altitude of the respective set and an altitude calculated from the temperature of the respective set and a vertical temperature gradient,

e. calculate, for the given moment, an altitude of the rain-snow limit for each type of air mass, each from the estimated average isothermal altitude for the type of air mass for the two instants immediately before the given instant for geographical portions in which precipitation is detected at the given moment. Other features, details and advantages of the invention will appear on reading the following detailed description, and the appended drawings, in which:

FIG. 1 shows the interactions of the components of a device according to the invention,

FIG. 2 represents an exemplary embodiment of a first function implemented by the device of FIG. 1,

FIG. 3 represents an exemplary embodiment of a second function implemented by the device of FIG. 1, and

- Figure 4 shows an embodiment of a third function implemented by the device of Figure 1. The drawings and the description below contain, essentially, elements of a certain character. They can therefore not only serve to better understand the present invention, but also contribute to its definition, if any. This description is likely to involve elements likely to be protected by copyright and / or copyright. The rights holder has no objection to the identical reproduction by anyone of this patent document or its description, as it appears in the official records. For the rest, he reserves his rights in full.

Reference is made to Figure 1. An isotherm forecasting device is referenced 1 as a whole. The device 1 comprises a memory 3, which is accessed by a sorter 5, a computer 7 and a comparator 9.

In the context of the invention, the memory 3 can be any type of data storage suitable for receiving digital data: hard disk, hard disk flash memory (SSD in English), flash memory in any form, RAM, disk magnetic, distributed storage locally or in the cloud, etc. The data calculated by the device can be stored on any type of memory similar to or on memory 3. This data can be erased after the device has completed its tasks or stored.

In the context of the invention, the sorter 5, the computer 7 and the comparator 9 are elements directly or indirectly accessing the memory 3. They can be made in the form of a suitable computer code executed on one or more processors . By processors, it must be understood any processor adapted to the calculation. Such a processor can be made in any known manner, in the form of a microprocessor for a personal computer, a dedicated FPGA or SoC chip ("System on chip"), a computing resource on a grid, a microcontroller, or any other form to provide the computational power necessary for the embodiment described below. One or more of these elements can also be realized in the form of specialized electronic circuits such as an ASIC. A combination of processor and electronic circuits can also be considered.

In the example described here, the memory 3 stores meteorological data in the form of sets associating: - a geographical portion i,

an altitude z of the measurement of the temperature on the ground,

- an instant h,

a floor temperature T,

- an altitude of the snow-rain limit: K or H,

- an altitude of the rain-snow limit between h and h + 1 IK or IH,

a type of air mass M, and

an indication of precipitation P, detected by the radar data and / or by the observation on the ground.

The memory 3 also stores one or more DTH coefficients in the form of vertical temperature gradients indicated. Indeed, it is considered here that in each geographic portion i, the temperature varies in the vertical direction while respecting a substantially constant coefficient. In the example described here, five DTH coefficients respectively worth 0.40 °; 0.65 °; 0.80 °; 0.95 ° and 1.00 ° for 100 meters of vertical drop from top to bottom. These values ensure that most temperature gradients are covered at low altitude. In the example described here, it is the DTH coefficient equal to 1.00 ° for 100 meters of elevation change from top to bottom is used. The index i makes it possible to identify each of the geographical portions. It takes the value of an integer between 1 and I, where I represents the total number of geographical portions managed by the device. The geographical area of work is thus divided into I geographical portions. The data representative of the geographical portions i take, here, the form of coordinates of the GPS type delimiting each geographical portion i. For example, the geographical area of work may be delimited by a square of about 150 kilometers of side while a mesh of the geographical zone defines geographic portions i square of one kilometer of side. In this example, the geographic area of work consists of 22500 geographical portions of the same shape, size and adjacent to each other. The index i therefore takes the values of integers between 1 and I, where I is equal to 22500. This example of the number of geographic portions i and the size of the geographical area is adapted to cover, for example, one of the plain regions around it. French Alps. The geographical area is inscribed in the cover of a radar (generally 200km x 200km) able to detect precipitation in a manner known in the state of the art. Alternatively, the shapes and dimensions may vary, in particular depending on the territory to be treated. For example, the geographical area may be delimited so as to be covered by several existing radars and / or one or more sensors on the ground. The dimensions and shapes of the geographical portions i may be heterogeneous within a geographical area, for example to improve the finesse of prediction on geographical portions presenting a specific meteorological behavior. In order not to estimate the evolution of the rain-snow boundary between two instants h with data from different geographical portions, the data are only used to estimate the altitude of the rain-snow boundary in the precipitating part of each mass. only at times h when all geographic portion data are updated. When the data of one of the geographical portions are updated with a greater frequency, these data will be used at each update only to determine the presence of the air mass on this geographical area at the corresponding times. As will appear in Figure 4, this increases the snow forecasting accuracy. Thus, in the context of FIG. 2, all the data are taken into account, and in the context of FIG. 3, only the data corresponding to times h for which the data are available for all the geographical portions are taken into account. .

The difference between the instant h and the forecast horizon h + 1 is chosen of constant duration, and corresponds to the very short term reliable forecasting deadline, that is to say with one hour to four ( never beyond) in advance of a phenomenon.

The temperature T of a set of data represents the temperature in the geographic portion i associated for a known altitude. This temperature T can be derived from a measurement by a local thermometer, generally on the ground, be derived from other types of measurements, for example satellite, or be deduced by the application of meteorological models known as such. The altitude of the isotherm IK or IH represents the altitude from which a reference temperature corresponding to the rain-snow limit is found in the associated geographical portion i. In other words, above this altitude, a lower temperature is expected, while below this altitude a higher temperature is expected.

This altitude is expressed, here, in meters and by reference to the sea level. In the example described here, the isothermal 1.5 ° C is chosen as the reference temperature because it is considered as representative of the low limit of the melting layer, that is the layer of the atmosphere which, when traversed by solid hydrometeors, allows at least 70% of its last to melt (see the article "Sensitivity of Precipitation Phase over the Swiss Alps to Different Meteorological Variables ", S. Froidurot et al, Journal of Hydrometeorology, April 2014, Vol.15, No.2: pp. 685-696).

As a result, precipitation that crosses this lower limit before reaching the ground has a liquid form: it is raining. On the contrary, when this low limit is near the ground or below ground level, then the precipitation does not melt before touching the ground: it snows. Of course, other parameters can in practice influence these weather phenomena. The melting may be partial (melted snow) or unstable (freezing rain). Depending on the phenomena that one wishes to predict and particular situations, the chosen value of the isotherm studied may be different from 1.5 ° C. The references K and H distinguish an altitude in a geographical portion i according to its association with a type of air mass M as described below.

One of the defects that handicap the devices of the prior art is that they treat all the geographically precipitating portions in isolation. The Applicant has discovered that gathering the precipitating geographical portions according to the type of air mass that occupies them at a given moment makes it possible to estimate more reliably the evolution of the altitude of the rain-snow limit in this air mass and thus to provide a forecast of this altitude.

The type of air mass M located on a geographical portion i at a time h is a boolean variable. In the example described here, each geographic portion i is assigned either an incoming air mass in the geographical area of work, or an air mass leaving the geographical area of work. When the geographical portion i is associated with an outgoing mass, the index M takes the value 0. When it is associated with an incoming mass, the index M takes the value 1. Thus, at each instant h, the geographical area can be decomposed into two sub-zones, each sub-zone being formed by the meeting of the geographical portions associated with an incoming mass, respectively outgoing. Meteorologically speaking, the incoming mass designates the mass of air behind an air front in the advancing direction of said front while the outgoing mass designates the mass situated in front of said front.

Because of the dimensions selected for the geographical area of the example described here, only one air front will be present. If a larger geographical area were chosen, it might be relevant to define several types of air masses, each associated with a front in that area.

The snow forecast SN at a point Q in a geographic portion i and at a time t is also a Boolean variable. It is attached to the presence of precipitation predicted by extrapolating the displacements of radar meteorological echoes. When the predicted precipitation is zero, SN is set to 0. Otherwise, the snow forecast will depend not only on the air mass in which the Q point is located, but also on the altitude of the point Q: if it is above the expected rain-snow limit, the SN index takes the value 1. The extrapolation forecast of radar weather echo movements is a known means using the comparison of the echoes between images provided by single or dual polarization radars with a frequency usually of 5 minutes. The reliability at the 70% threshold varies according to the forecasting deadline: it rarely exceeds 1h and is always zero beyond 4h. The precipitation forecast at a point Q is thus set to day every 5 minutes while the associated snow forecast is updated at each h hour of snow-to-snow limit assessment and to h + 1. Alternatively, the forecast P could be expressed as a percentage indicating the risk of precipitation.

Reference is now made to FIG. 2 which represents an example of implementation of a function by the sorter 5. The function of the sorter 5 is to assign one type of air mass to each geographical portion for each moment h. In an operation 500, the sorter 5 performs an initialization function Init (). The function Init () has for role to associate all the geographical portions with a mass of incoming air. Indeed, when the forecast starts, it is not possible to know a priori the distribution between incoming mass and outgoing mass, and it is logical to consider that all geographic portions are associated with an outgoing mass. Thus, the function Init () associates with each geographical portion i a type of air mass M equal to 0 for the moment h initial. The InitQ function also initialises two indices i and h with the value 1. These indices will be traversed sequentially, first in i, then in h, in order to subsequently assign to each geographical portion i a mass type of air for now h.

Then, in an operation 510, the sorter 5 interrogates the memory 3 with the pair of the current geographic portion i and the current instant h, in order to determine the absence of precipitation at time h. It should be noted that the geographical portions on which it does not rain and which are outgoing remain outgoing (until the resetting of operation 570). This is because, in the absence of precipitation, the phenomenon of absorption of the latent heat of the atmosphere at low altitude to allow the melting of the snow is not responsible for the altitude variation of the isotherms. As this phenomenon to which the invention is attached, it is not useful to detect a presence of different air mass that would not be related to a precipitation. Therefore, if there is an incoming air mass without precipitation, it is of no interest and remains marked outgoing. If the precipitation indication P for the moment h is non-zero, the sorter 5 executes an operation 520 in which it interrogates the memory 3 with the pair formed by the current geographical portion i and the current instant h to determine whether the type of air mass M associated with the geographical portion i is entering, that is to say if M has the value 1.

Indeed, because of the dimensions and phenomena chosen in the example described here, once a given geographical portion is associated with an incoming air mass, it is not necessary to change this association until that all other geographic portions are also associated with an incoming air mass. In other words, a geographic portion can change from the incoming air mass to the outgoing air mass only when the front has swept the entire geographical area.

If the air mass type M is outgoing (M has the value 0), then the sorter 5 will perform two tests to determine whether the associated air mass type needs to be changed. Thus, in an operation 530, the sorter 5 determines whether the temperature T of the geographical portion i between instants h-1 and h (separated by one hour in the example described here) has varied in absolute terms with a temperature. greater than a threshold value T1. If this is not the case, then in an operation 535, the sorter 5 determines whether the temperature T of the geographical portion i between the times h-2 and h (separated by two hours in the example described here) has varied. in an absolute manner a temperature greater than a threshold value T2. In the example described here, the threshold value T1 is chosen equal to 1.5 ° C., and the threshold value T2 is chosen equal to 2 ° C. In both cases, this represents a significant change in temperature over a period of less than two hours.

If one of the operations 530 or 535 detects a threshold overrun, then the identified temperature change is considered to indicate that the edge has passed over the geographical portion in question, and that the type of air mass associated with it must be modified. This is done in an operation 540.

If one of the operations 510 and 520 is negative, or both operations 530 and 535 are negative, then the type of air mass must not be changed. In this case, or after operation 540, a test is performed in an operation 550 to determine if all geographical portions have been traveled for the moment h. If this is not the case, then the index i is incremented in an operation 555 to move to the next geographical portion, and the loop resumes with the operation 510.

If all the geographical portions have been traveled for the moment h, then, in an operation 560, the sorter 5 determines whether all the geographical portions are associated with an incoming air mass or if it no longer rains on any geographical portion. by executing an Rst () function. If this is the case, then in an operation 565, a function Reinit () operates in the same way as the function InitQ, but without resetting the indices i and h.

Then, or when the test of the operation 560 is negative, the index h is incremented in an operation 570, then a test in an operation 580 determines whether all times h have been traveled. If this is not the case, the index i is reset to designate the first geographical portion in an operation 585, and the loop resumes with the operation 510. Otherwise, the function ends in an operation 590.

In the example described here, all the temperature data are obtained at identical times for all the geographical portions. Thus, all geographical portions are traversed by the operation loop 510 to 560, then the index h is incremented. As described above, when certain geographical portions present updated data more frequently than all hours, the operations 510 to 560 can be performed for the corresponding instants, in order to know more precisely the membership of these geographical portions to an incoming or outgoing mass.

This makes it possible to be more precise at the time of the snow forecast as described with FIG. 4. In this case, the function of FIG. 2 can be modified to invert the course of the indices i and h: all the indices h are covered for a given geographical portion i, then the index i is incremented. At the end of FIG. 2, only the moments h for which data are available for all the geographical portions are used for the execution of the function described with FIG. 3. Reference is now made to FIG. which represents an example of implementation of a function by the computer 7. The purpose of the computer 7 is to determine the isothermal altitude for the incoming and outgoing air masses from the data calculated by the sorter 5 The computer starts, in an operation 700, indices i and h initialized with the value 1. The function of FIG. In a first part, isothermal altitudes at the reference temperature are calculated for each geographic portion associated with a precipitation, as a function of a DTH coefficient. Once all isothermal altitudes at the reference temperature have been calculated with the most relevant DTH coefficient. Then, in a second part, the isothermal altitudes of the reference temperature for the first and second types of air mass are calculated.

Thus, in an operation 705, the computer 7 interrogates the memory 3 with the pair formed by the geographical portion i and the instant h current, and determines whether the precipitation indication P associated therewith indicates the absence of precipitation.

If, on the contrary, the precipitation prediction P indicates a precipitation, that is to say if P has the value 1, then, in an operation 710, the computer 7 interrogates the memory in order to determine which isothermal altitude elementary will be calculated. For this purpose, the type of air mass M associated with the pair formed by the current geographical portion i and the current instant h is tested, for example by testing whether M has the value 1, which corresponds to an air mass. entering. If this is the case, then in an operation 715, the computer 7 calls the memory 3 with the data pair associating the geographical portion i and the instant h current. It extracts a temperature T in the geographical portion i, at the instant h, and than an altitude z at which this temperature has been determined. The temperature T results, for example, from a ground measurement carried out at a known altitude.

The calculator 7 also obtains the coefficient DTH, and calculates accordingly the elementary isothermal altitude of the incoming air mass K (i, h) for the geographical portion i at the instant h and the coefficient DTH by applying the formula of operation 715, in which Tiso is the temperature for the isotherm corresponding to the reference temperature. This formula calculates the altitude of the Tiso isotherm by adding to the altitude at which the temperature T has been measured a height corresponding to the difference between the temperature T and the isothermal temperature Tiso, multiplied by the temperature gradient DTH . In the example described here, it is desired to locate the melting temperature of hydrometeors. For this, the temperature of the Tiso isotherm, referred to as the reference temperature, is chosen to be 1.5 ° C.

If the type of air mass M is outgoing, then an operation 720 is executed. In this operation, the formula is identical to that of operation 715, but it is a reference isothermal altitude of outgoing air mass H (i, h) which is calculated. This is important because isotherms of incoming air mass and outgoing air mass are treated separately depending on the type of air mass. Nevertheless, as a variant, the operations 710 to 720 could be carried out in a different manner, by performing the calculation of the formula of the operations 715 and 720 and by assigning a label at the resulting element isothermal altitude according to the type of mass of the air, subsequent calculations can be based on this label. Then, or when the operation 705 indicates that no precipitation is expected for the geographical portion i at time h, a test is performed in an operation 725 to determine if the reference isothermal altitudes of all the portions. geographical have been determined for the moment h. If this is not the case, then the index i is incremented in an operation 730 to determine the reference isothermal altitudes of the next geographical portion at the instant h, then the first part resumes with the operation 705 .

Once the reference isothermal altitudes have been calculated for all geographic portions at time h, the second part can be executed.

For this, two isothermal elevations K (h) and H (h) are calculated in an operation 735 by means of an Avg () function. The Avg () function uses the elementary isothermal altitude values K and H calculated with the first loop and calculates their respective mean. Thus K (h) is the mean over i of K (i, h) and H (h) is the mean over i of H (i, h). The calculator 7 then implements an operation 740, in which the value of the isothermal altitude IK (h + 1), for all the geographical portions i, between the instant h and the moment h + 1 succeeding it . It is considered that the altitude of the isotherm IK (h + 1) will evolve linearly with respect to the previous moment. Thus, the isotherm IK (h) is calculated as the sum of K (h) and the linearization of the evolution of K (h) over time, ie half of the difference between K (h) and K (hl).

In the same way, the calculator 7 determines in an operation 745 the value of the isothermal IH (h + 1) as the sum of H (h) and the linearization of the evolution of H (h) in time, that is, the difference between H (h) and H (hl) divided by two (because the difference is taken over two units of time).

Alternatively, operations 740 and 755 may be performed in reverse order or in parallel. Then, the computer 7 executes an optional operation 750. In this operation, the computer 7 executes an Exc () function. The Exc () function corrects some exceptional cases. For example, the isotherm may evolve more rapidly for the incoming air mass than the prediction is able to calculate. Thus, for the incoming air mass, the temperature of each of the associated geographical portions is determined, and if one or more of these portions has a temperature lower than Tiso (1.5 ° C. in the example described), is the lowest altitude associated with each of these measurements that is selected as IK (h + 1). Indeed, since the Tiso temperature (or a lower temperature) has already been observed, there is no sense in defining an isothermal altitude IK (h + 1) which is more important. The calculator 7 then tests the index h to see if all the predictions have been made in an operation 755. If it is not the case, then the index h is incremented in an operation 760, and the index i is reset to 1 in an operation 765.

If all the isothermal forecasts have been made for all times h, then the function ends in an operation 790.

After these operations, the memory 3 stores meteorological data sets in which, for each pair of geographic portion i and moment h, an isothermal altitude IK or IH is associated. This isothermal altitude IK or IH can be used for various weather purposes. In the following, the altitude of the reference isotherm, here isotherm 1.5 ° C, is used to predict a snowfall.

Reference is now made to FIG. 4 which represents an example of implementation of a function by the comparator 9. In this function, the comparator 9 determines a snow forecast for a location Q at a given instant t.

In an operation 900, the comparator 9 performs an Init () function. The function InitQ determines the coordinates of the point Q, as well as the index h which corresponds to the desired time of prediction.

Then, in an operation 910, the comparator 9 executes a PG () function which determines the geographical portion to which the Q point is associated. The comparator 9 executes an operation 920, in which a function Prev () receives as its argument a pair formed with the geographical portion i and the current instant t, and determines whether the precipitation prediction indicates the absence of precipitation. In the example described here, the memory 3 receives all precipitation forecasts for each geographical portion within each time slot. Alternatively, these forecasts could be the subject of a specific request from a server.

If precipitation is predicted, that is, if P has the value 1, then the nearest instant h preceding the instant t for which a snow-to-snow limit altitude has been calculated is determined. Then, in an operation 930, the comparator 9 executes a function SN () which receives the altitude z from the point Q and the values i and h as arguments. If no precipitation is expected, then the function ends. The function SN () has the function of determining a snow forecast according to the variables i and t. For this, the comparator 9 determines whether the geographical portion i is associated with an incoming or outgoing air mass.

If the geographical portion i is associated with an incoming air mass, then the altitude z of the point Q and the isothermal altitude IK (h) are compared If z is less than IK (h) then SN () indicates an absence of snow forecast. If z is greater than IK (h) then SN () indicates a snow forecast.

If the geographical portion i is associated with an outgoing air mass, then the altitude z of the point Q and the isothermal altitude IH (h) are compared. If z is less than IH (h) then SN () indicates no snow forecast. If z is greater than IH (h) then SN () indicates a snow forecast.

Then the function ends in an operation 940 with the return of the snow forecast (or its absence). The function shown in Figure 4 is quite simple, and returns a forecast for a given location. It is nevertheless easy to adapt it to more complex and / or rich situations. For example, as precipitation forecasts may be more frequent (such as a five-minute forecast) than hours (which are usually separated by one hour), forecasts can be updated. snow for a given location every five minutes. Similarly, when a route is considered, it is possible to provide snow forecasts along the route by establishing the snow forecasts along the route at the times at which a driver is expected to pass each point concerned. In addition, these forecasts can be communicated in various ways and in particular on a form compliant with the patent FR 2 864 850, or any other device for ergonomically displaying such forecasts.

In the above, the forecasts have been presented as Boolean data. In variants, the calculations could be done differently to return a value representative of a probability instead of a Boolean value.

The invention is not limited to the examples of isothermal prediction device described above, only by way of example, but it encompasses all the variants that may be considered by those skilled in the art within the scope of the present claims. -after.

Claims

claims

1. Apparatus (1) for predicting altitude of the rain-snow boundary comprising

a memory (3) for receiving sets of meteorological data, each set of meteorological data associating a geographical portion (i), an instant (h), a temperature (T), a precipitation indication (P), at least one altitude (K, H, IK, IH), and a type of air mass (M) indicating whether an air mass associated with the geographical portion (i) for a moment (h) is considered as entering or leaving a geographical area corresponding to the union of the geographical portions (i), each set of meteorological data being initialized with a type of air mass (M) indicating an outgoing air mass, and the instants (h) being chosen so that meteorological datasets can be grouped by moment (h) for a given geographic portion (i),

a sorter (5) arranged to traverse a group of meteorological data sets comprising a current instant (h), and, for a given set of this group, to modify the type of air mass (M) to indicate a mass of incoming air when the difference between, on the one hand, the temperature (T) of the given set and, on the other hand, the temperature (T) of a set associating the geographical portion (i) of the given set with a instant (h-2; h-1) prior to instant (h) current is greater than a threshold value (T1, T2),

a calculator (7) arranged to calculate, at a given moment (h), a predicted altitude of the rain-snow limit (IK, IH) for each type of air mass (M), each starting from a mean isothermal altitude (K, H) estimated for the respective air mass type (M) for the two instants immediately preceding the given instant (h) for geographical portions in which precipitation is detected at the given instant (h), which isothermal altitude (K, H) estimated for the given instant (h) and a given type of air mass (M) is calculated from the mean for the data sets including the instant (h) and type of air mass (M) given, the sum of the altitude (z) of the respective set and an altitude calculated from the temperature (T) of the respective set and a vertical temperature gradient (DTH).

2. Device according to claim 1, wherein the altitude calculated from the temperature (T) of the respective set and a vertical temperature gradient (DTH) corresponds to the product of the vertical temperature gradient (DTH) by the difference between the temperature (T) of the respective set and the isothermal temperature whose altitude corresponds to that of the rain-snow limit.

3. Device (1) according to claim 1 or 2, wherein the sorter (5) is arranged to determine whether the type of air mass (M) must be modified by comparing the temperature (T) of the given set and on the other hand the temperature (T) of a set associating the geographical portion (i) of the given set with the instant (h-1) immediately preceding the current instant (h).

4. Device (1) according to one of the preceding claims, wherein the sorter (5) is arranged to determine whether the type of air mass (M) must be modified by comparing the temperature (T) of the assembly given and on the other hand the temperature (T) of a set associating the geographical portion (i) of the given set with the second instant (h-2) immediately preceding the current instant (h).

5. Device (1) according to one of the preceding claims, wherein the sorter (5) is further arranged to change the type of air mass (M) to indicate a mass of air leaving meteorological data sets comprising a given instant, when all data sets comprising the instant immediately preceding said given instant have a type of air mass (M) indicating an incoming air mass.

6. Device (1) according to one of the preceding claims, wherein the isothermal altitude (IK, IH) estimated for each type of air mass (M) results from a linearization of the evolution of values. of average isothermal altitude at the previous instants.

7. Device (1) according to one of the preceding claims, wherein the computer (7) is further arranged to repeat the calculations, for times (h + 1; h + 2) after the moment (h) current, after the calculations for the moment (h) current.

8. Device (1) according to one of the preceding claims, wherein the computer (7) is further arranged to correct the isothermal altitude (IK, IH) estimated for a given instant (h) for the type of incoming air mass, when a temperature (T) below the isothermal temperature whose altitude is sought is detected at the given instant (h).

9. Device (1) according to one of the preceding claims, further comprising a comparator (9) arranged to compare the isothermal altitude (IK, IH) estimated for a moment (h) and a geographical portion (i) at a ground elevation (z) of a location of said geographic portion (i), and inferring a snowfall prediction at said location at time (h) when said elevation (z) is higher.

10. Rain-snow limit elevation prediction method comprising

at. receive sets of meteorological data, each set of meteorological data associating a geographical portion (i), a time (h), a temperature (T), an indication of the presence of precipitation (P), at least one altitude (K, H , IK, IH), and a type of air mass (M) indicating whether an air mass associated with the geographical portion (i) for a moment (h) is considered as entering or leaving a corresponding geographical area at the union of the geographical portions (i), the instants (h) being chosen in succession, so that the sets of meteorological data can be grouped by instant (h) for a given geographical portion (i),

b. initialize each set of meteorological data with a type of air mass (M) indicating an outgoing air mass,

vs. traversing a group of meteorological data sets including a current instant (h), and, for a given set of this group, changing the type of air mass (M) to indicate an incoming air mass when the difference between on the one hand the temperature (T) of the given set and on the other hand the temperature (T) of a set associating the geographical portion (i) of the given set and an instant (h-2; h-1) prior to the instant (h) current is greater than a threshold value (T1, T2),

d. calculating an estimated isothermal altitude (K, H) for an instant (h) and an air mass type (M) given from the mean, for the meteorological data sets including the instant (h) and the type of air mass (M) given, the sum of the altitude (z) of the respective set and an altitude calculated from the temperature (T) of the respective set and a vertical temperature gradient (DTH),

e. calculating, for the moment (h) given, an altitude of the rain-snow limit (IK, IH) for each type of air mass (M), each starting from the average isothermal altitude (K, H) estimated for the type of air mass (M) respectively for the two instants immediately preceding the given moment (h) for geographical portions in which precipitation is detected at the given instant (h).