I see that your reference plot x axis is bending angle change and not just bending angle itself
so I wonder if you should not take the diff or gradient of this angle
I get a new plot that shows better the bump around altitude = 33 km but at lower altitudes I see still a large discrepancy between your data and the reference plot - maybe it's not related to how we process it in matlab but the nc file itself ?
File = "atmPrf_C2E3.2022.016.01.45.R04_0001.0001_nc";
info = ncinfo(File)
impactHeigth = ncread(File,"Impact_height");
Temp = ncread(File,'Ref');
tongLat = ncread(File, "Lat");
TongLong = ncread(File, "Lon");
pres = ncread(File,"Pres");
bendAngle = ncread(File, "Bend_ang");
bendAngle_sigma = ncread(File, 'Bend_ang_stdv');
bendAngle_ci = ncread(File, 'Bend_ang_conf');
MSL = ncread(File, "MSL_alt");
maxHeight = 40;
minHeight = 5;
alt = MSL((MSL >= minHeight) & (MSL <= maxHeight));
temp = Temp(1:length(alt));
figure
plot(bendAngle, impactHeigth)
ylim([10 max(impactHeigth)]);
title("COSMIC-2 RO Bending Angle W.R.T Mean Sea Altitude")
xlabel("Bending Angle")
ylabel("Mean Sea Altitude (km)")
figure
plot(gradient(bendAngle), impactHeigth)
ylim([10 max(impactHeigth)]);
title("COSMIC-2 RO Bending Angle W.R.T Mean Sea Altitude")
xlabel("Gradient of Bending Angle")
ylabel("Mean Sea Altitude (km)")
