The first thing to notice on that chart is the actual voltage values. The Arduino is not going to like those negative voltages.
The second thing to notice is that there is no horizontal scale identification. Are those pulses 20 nanoseconds apart or two weeks?
The third thing you would need to know is what was being transmitted during that measurement period. Is that temperature data only? Temp and humidity? What was the temperature?
Notice 2 bytes of 0xFF, think these are markers or separators to get the receiver in sync or so.
leaving 2 values { 75-80-138 } and { 87 }
assumption 1: the single byte is humidity - 87% would be a nice reading
assumption 2: the triplet is the temperature, three values 2 with first bit 0 and 1 with first bit set
assumption 3: The one with the first bit set is the fraction part and the other 2 are the decimal part.
So we have { 75, 80 } . {138}
If we "remove" the decimal point bit we are left with {75, 80} . {10} But I have no idea what value was on the receiver or how to translate it further......
I suspect that the graph i see is inverted, so I will try that as well.
Also, probably the most significant bit of the temperature is reserved for negative values.
silly from me, i didn't write up the values reading on the screen when recording the signal.. but I recall it was something around 10-14 degrees Centigrade as I just brought it indoors. will repeat that later.
This sensor only sends temperature. I suspect the last byte is a CRC
try sampling more values, with double figures or triple, e.g.
22.0
20.2
22.2
This would make it easier to figure out a pettern.
My Oregon Scientific temperature sensor gives this signal for 30.0 °C:
But I don't think your sensor uses same coding: manchester coding can be quickly indetified as you never see three consecutive identical figures (111, or 000), while your picture appears to be something like:
01101101101101101101101100011011000110001101100011011011011011000110001100011000110001100011011011011011000110001100011000
Bold sequences (I didn't highlight them all) can't exist in manchester coding, made of sequences of "01" and "01".
Anyway, looking at your two images, I think I can see the "preamble", used to "tune up" the receiver, the same in both images:
01101101101101101101101100011011000110001101100011011011011011000