"Last" Doesn't Always Mean "Previous" - 4b

Graph One

Computer languages not unlike human languages have nomenclature problems from small to large.

The title of this series brings to mind a small one, while this link brings to mind a deadly one.

This series is about software that analyzes the atmosphere using analysis techniques originally revealed by Josiah Gibbs.

The exhaustive coverage of the relevant software in Fortran and Visual Basic can be studied here. In my struggle to convert the Fortran version into C++ I have moved along enough to generate some graphs using the new C++ version (Graphs One thru Three)..

Graph Two

The mind-blowing science at work digs into the measurements of the atmosphere the way the TEOS-10 C++ software did while focusing on measurements of the ocean depths (World Ocean Database).

The global average atmospheric temperature anomalies (not ocean) can be viewed for Berkeley, NOAA, and GISSTEMP, 

The graphs today use the Berkeley anomaly values. The CSV data for generating the graphs begins with a 2024 global climate temperature average in degrees Kelvin (K) as follows:

Global Average Temperature (2024) = 288.33 K
°C = K - 273.15
°C = 288.33 - 273.15
°C = 15.18 [NASA says 15.10]

°F = (°C)15.18 * (9/5) + 32
°F = 59.324

On 22 July 2024, the daily global average
temperature reached a new record high of
17.16°C (62.888 °F) (351.218 °K).

(See: Copernicus, WMO, NASA). On with the show ...

Graph Three

There is a seemingly strange characteristic that is noticeable in the graphs (Graph Three shows a decrease while Graph One and Graph Two show an increase).

This is a recognized phenomenon in the science : "If the gas is truly ideal, the specific heat capacity is temperature-independent" (Properties of Common Gases/Steam and Moist Air with Temperature). This variability is a current base of professional argument discussed by Dr. James Hansen in the video presented in ("Last" Doesn't Always Mean "Previous" - 2).

Dr. Hansen argues that the software models do not properly recognize internal dynamics of the impact of varying mixes of atmospheric gasses (nitrogen, oxygen, trace amounts of argon, carbon dioxide, methane, water vapor,, ozone,, and other gases like neon, helium, and krypton),

Notice that in the three graphs above the temperature and density data are the same, however, as it turns out the results returned by the same function are a feature, not a bug:

"Climate change caused by human activities is influencing atmospheric pressure, according to a new study. The research, published in this week’s Nature, is the first report of a human-induced effect on global climate that does not rely on measurements of temperature.

Nathan Gillett of the University of Victoria in Canada and colleagues found that air pressure has decreased on average over the Arctic, Antarctic and North Pacific during the past five decades. In contrast, above the North Atlantic, southern Europe and North Africa, pressure has increased.

These changes in atmospheric pressure could have important consequences for future climate, owing to influences on patterns of rainfall, air temperatures, winds and storminess. This means that current climate predictions — which currently fail to take account of regional effects of air pressure changes — may be unreliable."

(SicDev). It boils down to: "Uncertainties in projected warming stem from three sources: uncertainty in future emissions, uncertainty in the climate response to those emissions, and internal variability" (Nature).

 The just in case paths inside the function are like the gasses in the atmosphere::

  double tr,dr,ra,f,tau,del,a,a_t,a_tt,a_d,a_td,a_dd;

  ra = gas_constant_air_L2000; [SPECIFIC GAS CONSTANT OF AIR IN J                 KG-1 K-1 USED BY LEMON ET AL. 2000]
    

tr = 132.6312;

dr = 10447.7 * ma; 

 [ma = molar_mass_air_L2000   (MOLAR MASS OF AIR IN KG/MOL USED BY     LEMMON et al. 2000]

  init_iapws10();

 tau = tr / t_si;
 del = d_si / dr;

case (drv_t 1, drv_d 1):
                    a_d = 1 / del + alpha_res(0, 1, tau, del);
                    a_td = alpha_res(1, 1, tau, del);
                    f = ra * (a_d - tau * a_td) / dr;
       
case (drv_t 0, drv_d 1)::
                    a_d = 1 / del + alpha_res(0, 1, tau, del);
                    f = ra * t_si * a_d / dr;

case (drv_t 2, drv_d 0)::
                    a_tt = alpha_ideal_tt(tau) + alpha_res(2, 0, tau, del);
                    f = ra * pow(tau,2) * a_tt / t_si;

case (drv_t 0, drv_d :
                    a_dd = -1 / pow(del,2) + alpha_res(0, 2, tau, del);
                    f = ra * t_si * a_dd / pow(dr,2);

The variation in the graphs stems from the value of function parameters "drv_t" and "drv_d", because the other two parameters ("t_si" and "d_si") are the same in all three graphs.

Closing Comments

The d_si values are averages of atmospheric pressure from 1850 - 2024 recorded in over 404,000 weather stations around the globe; the t_si values are averages of global temperatures per Berkeley anomaly values.

Data is available for download:  Lundstad, Elin; Brugnara, Yuri; Brönnimann, Stefan (2022): Early instrumental time series of global measurements of monthly precipitation between 1677-2021, part 1 [dataset]. PANGAEA, (2022): Global Early Instrumental Monthly Meteorological Multivariable Database (HCLIM) [dataset bundled publication]. PANGAEA).

This series is about TRENDS ... the trends of change that can be tricky and cause the best scientists to miss the bus, miss the train, miss the boat, and miss the mark (see video below).

The previous post in this series is here.

---

---