searching for high precision altimeters

Hello everyone, I am creating small high-altitude stations net to be put on a glacier at 3.600 meters, to record the gradual snow melting.
The snowpack decrease on average by 7 meters, between march and the summer.
I'm using for the moment these altimeters:
https://learn.sparkfun.com/tutorials/mpl3115a2-pressure-sensor-hookup-guide
But also in my house the measurement reveal a 2 meters fluctuation, not 0.3 meter as said.
Obviously on the mountain we will also record a pressure at a known point and not on the snow, but I would ask:
You know the most accurate and professional sensors?

Thank you all

An cheap one is BMP180 (more members in that family)
Specs are good. See the pdf

I read differece when rising/lowering the sensor by 30cm.
(when comparing the two: can datasheets be trusted?)

I've used a number of pressure sensors, and the best in my experience is the LPS25H (least noise and highest precision for the price). Average at least 100 readings for best performance.

This seems to be a good solution, and what if the budget is even higher?
There is nothing between 30 and 50$ that really make the difference?

However accurate the sensors are , barometric pressure is continuously varying even for a stationary sensor.

Using a differential sensor on a tube with one end of the tube mounted at the reference height may give a better indication.

Sorry, can you better explain the tube concept?
Anyway, we are planning already to put a sensot on a know height point, and using it as a reference, but i'm not sure is what you mean talking about the tube.

Rather than use two sensors, use one differential one with a tube off each side.

One tube end at fixed reference height other tube end at measure height.

Then just measure the differential pressure rather than the absolute.

In this way local pressure variations should cancel out.

That is the theory anyway , ill leave it to others to comment on.

EDIT
thinking about it though, there is still the mass of air in the tubes so i am not at all sure this is a valid idea.

There is nothing between 30 and 50$ that really make the difference?

Probably, but this is a hobby forum and many people don't want to spend more than $5 on anything, so you are unlikely to get expert advice on them.

Look carefully at the sensor specification sheet. In particular you want to study the "rms noise" (pay attention to the units), which will give a good idea of the maximum sensitivity and stability. It will also depend on any internal averaging that the sensor performs. For example, the LPS331AP sensor has rms noise = 0.45 mBar for no averaging, and 0.02 mBar for maximum averaging.

0.02 mBar pressure change corresponds to a change in altitude of about 17 cm at sea level.

Having two sensors (one fixed) is a fine idea. However, when the wind blows you can get surprisingly large pressure variations over a very short distance, especially on slopes. This will be obvious when you start looking at the data.

The combined noise of two sensors is sqrt(2) times the noise of one.

Actually the idea is to have 25 sensors (inside protective cases) on a snowfield and a reference sensor (exactly the same case and electronic) inside a mountain hut, 3km far away.
Of course i can take several samples on each sensors and do an average, but i'm wondering myself if 3km is not an excessive distance.
I'm having kind of a 15meters fluctuation today, with the sensor in my house.
Actually i'm asking the module to give me the altitude, a better idea would be to just ask him the pressure and do all the calculation on serverside when i receive the data?
From the sparkfun website:

It would be easier to do all the calculations at the base station, as you can do post-processing and change the programs much more easily. Be sure to collect temperature data for each sensor.

Pressure differences due to wind will be a serious problem over 3 km distance, especially if there is a prevailing wind direction, but long term averaging and other forms of data processing should help. You will definitely need an anemometer and wind direction indicator at the base station.

How do you plan to synchronize all the measurements?

I did not thinked to the anemomether but is a nice idea, even if there is a moutain between the two points.

This is the snowfield:

This is the hut:

I think to sync the system by putting a RTC inside each module.
Each module will have three temperature sensors: one inside the altimeter, one outside, one in the battery pack, to understand when to heat them to allow the solar panel to recharge.

jremington:
I've used a number of pressure sensors, and the best in my experience is the LPS25H (least noise and highest precision for the price). Average at least 100 readings for best performance.

I've just got the sensor.
Do you average yourself the readings or there is library function?
Some altimeter handle this themself.

Day to day changes in atmospheric pressure due to weather changes can lead to hundreds of meters difference in the apparent altitude.

You don’t need a library to average: just four lines of code:

long sum=0;
for (int i=0; i<100; i++) {
   sum += read_sensor();
   }
sum = sum/100;