Models for Tropical Climate Dynamics

Waves Clouds and Precipitation

Boualem Khouider
Visit to download the full and correct content document:
https://textbookfull.com/product/models-for-tropical-climate-dynamics-waves-clouds-a
nd-precipitation-boualem-khouider/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

An Introduction to Clouds From the Microscale to

Climate 1st Edition Ulrike Lohmann

https://textbookfull.com/product/an-introduction-to-clouds-from-
the-microscale-to-climate-1st-edition-ulrike-lohmann/

An Introduction to Clouds From the Microscale to

Climate 1st Edition Ulrike Lohmann

https://textbookfull.com/product/an-introduction-to-clouds-from-
the-microscale-to-climate-1st-edition-ulrike-lohmann-2/

Mathematical Models for Neglected Tropical Diseases

Essential Tools for Control and Elimination Part B 1st
Edition Maria Gloria Basáñez And Roy M. Anderson (Eds.)

https://textbookfull.com/product/mathematical-models-for-
neglected-tropical-diseases-essential-tools-for-control-and-
elimination-part-b-1st-edition-maria-gloria-basanez-and-roy-m-
anderson-eds/

Demystifying Climate Models A Users Guide to Earth

System Models 1st Edition Andrew Gettelman

https://textbookfull.com/product/demystifying-climate-models-a-
users-guide-to-earth-system-models-1st-edition-andrew-gettelman/
Carbon Management in Tropical and Sub Tropical
Terrestrial Systems Probir K. Ghosh

https://textbookfull.com/product/carbon-management-in-tropical-
and-sub-tropical-terrestrial-systems-probir-k-ghosh/

Supercritical Antisolvent Precipitation Process

Fundamentals Applications and Perspectives Diego T.
Santos

https://textbookfull.com/product/supercritical-antisolvent-
precipitation-process-fundamentals-applications-and-perspectives-
diego-t-santos/

Dynamics and Predictability of Large Scale High Impact

Weather and Climate Events 1st Edition Jianping Li

https://textbookfull.com/product/dynamics-and-predictability-of-
large-scale-high-impact-weather-and-climate-events-1st-edition-
jianping-li/

Computational Methods for Nanoscale Applications

Particles Plasmons and Waves Igor Tsukerman

https://textbookfull.com/product/computational-methods-for-
nanoscale-applications-particles-plasmons-and-waves-igor-
tsukerman/

Climate Change and Energy Dynamics in the Middle East

Modeling and Simulation Based Solutions Hassan Qudrat-
Ullah

https://textbookfull.com/product/climate-change-and-energy-
dynamics-in-the-middle-east-modeling-and-simulation-based-
solutions-hassan-qudrat-ullah/
Mathematics of Planet Earth Series 3

Boualem Khouider

Models for
Tropical Climate
Dynamics
Waves, Clouds, and Precipitation
Mathematics of Planet Earth

Volume 3

Series editors
Ken Golden, The University of Utah, USA
Mark Lewis, University of Alberta, Canada
Yasumasa Nishiura, Tohoku University, Japan
Joseph Tribbia, National Center for Atmospheric Research, USA
Jorge Passamani Zubelli, Instituto de Matemática Pura e Aplicada, Brazil
Springer’s Mathematics of Planet Earth collection provides a variety of well-written
books of a variety of levels and styles, highlighting the fundamental role played
by mathematics in a huge range of planetary contexts on a global scale. Climate,
ecology, sustainability, public health, diseases and epidemics, management of
resources and risk analysis are important elements. The mathematical sciences play
a key role in these and many other processes relevant to Planet Earth, both as a
fundamental discipline and as a key component of cross-disciplinary research. This
creates the need, both in education and research, for books that are introductory to
and abreast of these developments.

More information about this series at http://www.springer.com/series/13771

Boualem Khouider

Models for Tropical Climate

Dynamics
Waves, Clouds, and Precipitation

123
Boualem Khouider
Mathematics and Statistics
University of Victoria
Victoria, BC, Canada

ISSN 2524-4264 ISSN 2524-4272 (electronic)

Mathematics of Planet Earth
ISBN 978-3-030-17774-4 ISBN 978-3-030-17775-1 (eBook)
https://doi.org/10.1007/978-3-030-17775-1

Mathematics Subject Classification (2010): 34Lxx, 35-xx, 35Pxx, 35Q30, 35Q31, 35Q40, 45Jxx, 60Jxx,
62F15, 65Cxx, 65M06, 15A16, 15A18

© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To my wife, Aldjia, who supported and
tolerated my long hours in the office. To my
children, Ines and Mohand. To the memory of
my parents.
Preface

Mathematicians and physicists are no strangers to the challenging problems of the

century, posed by global warming and climate change. It is even more meaningful
for them to get immersed into the core of climate modeling and into improving the
tools that are used for day-to-day weather forecasting and climate predictions. While
enormous progress has been made in terms of improving the fidelity of these large
computer models to represent the climate system, climate models are still strug-
gling in accurately representing the mean and variability of rainfall and wind pat-
terns associated with waves and clouds in the tropics. This volume reviews a series
of mathematical models for tropical convection and convectively coupled waves,
developed recently by the author and his collaborators. It blends in physical intu-
ition, gained from observations, with the state-of-the-art mathematical tools such as
Galerkin truncation, linear stability analysis, numerical methods, and stochastic pro-
cesses to capture and understand the subtle physics of wave-convection interactions
and cloud dynamics. It is intended for graduate students in applied mathematics and
physics interested in climate modeling and tropical meteorology, in particular. It as-
sumes prior knowledge of basic theory of partial differential equations, numerical
analysis, and fluid mechanics. Exposure to meteorology and stochastic processes
can be helpful but not necessary.
With the warming climate threat, it is of vital importance for the human kind to
keep improving the only tools we have in order to understand climate change and
make faithful forecasts. Despite the world-wide coordinated efforts and subsequent
improvements in our climate modeling capabilities, during the last decades, current
climate models suffer from various biases that make them less reliable in terms of
predicting local changes in the climate system, needed for decision-making at the
regional and the continental level. The long-range predictions of natural phenom-
ena, such as major droughts and floods, or the intensity and frequency of tropical
cyclones remain very uncertain, especially in some of the highly populated tropical
regions such as India and the Philippines. There is strong scientific evidence that
these biases are associated with the way in which climate models treat physical pro-
cesses associated with tropical precipitation and clouds. Progress in this regard is
tied to a better understanding of these processes and how the associated unresolved
vii
viii Preface

convective motions interact with synoptic and planetary-scale waves. Exposing this
subject to the wider audience, including mathematicians and physicists, will not
only increase our chances of getting there much quicker. Moreover, organized tropi-
cal convection and convectively coupled tropical waves offer many challenging sci-
entific and mathematical problems that deserve attention from some of the bright-
est minds. They involve a wide spectrum of physical phenomena ranging from the
phase change of water occurring at the micro-scale to the role of the change in sign
of the Coriolis force at the equator that drives very unique wave phenomena oc-
curring at various scales and to the direct interactions of clouds with atmospheric
dynamics and the climate system as a whole. Various teleconnections patterns vary-
ing on weekly to monthly time scales are known to exist between tropical wave
dynamics and mid to high-latitude weather and climate variability. As such, a good
understanding and better representation by climate models of these waves are very
important for weather predictions all over the globe on multiple time scales ranging
from days to seasons.

Victoria, BC, Canada Boualem Khouider

Acknowledgments

This volume and the contained research wouldn’t have been possible without the in-
volvement of many collaborators, students, and post-docs. I am especially grateful
to my longtime collaborator and post-doc mentor, Prof. Andrew Majda, who made
me discover this fascinating subject. I have been nourished through the years by
being able to interact and occasionally collaborate with some of the most famous
experts in the field such as George Kiladis, Mitch Moncrieff, and Joseph Tribbia,
I was standing on the shoulders of giants. Special thanks to the Banff International
Research Station for giving me the opportunity to organize workshops on these top-
ics which allowed me to broaden my horizons and interact with many people. The
Pacific Institute for the Mathematical Sciences CRG on mathematical problems in
climate modeling was one of the precursors for the success of this research. I ac-
knowledge the generous support from the Natural Sciences and Engineering Re-
search Council of Canada, the Canadian Foundation for Climate and Atmospheric
Research, and the Indian Institute for Tropical Meteorology, without which, many
projects in this research won’t have been even started.

ix
Contents

Part I Background and Preliminaries

1 The Governing Equations and Dry Dynamics . . . . . . . . . . . . . . . . . . . 3

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 The Primitive Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Equatorially Trapped Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.1 The Shallow Water Approximation . . . . . . . . . . . . . . . . . . . 8
1.4 The Vertical Normal Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.5 Riemann Invariants and Meridional Expansion . . . . . . . . . . . . . . . . 18
1.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2 Moisture and Moist Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.1 Moisture Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.1 Equation of State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.2 Cloudy and Precipitating Atmosphere . . . . . . . . . . . . . . . . . 25
2.2 The First Law, the Second Law, and Other Thermodynamic
Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3 Phase Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 Conserved Moist Thermodynamic Variables . . . . . . . . . . . . . . . . . . 30
2.5 Processes Leading to Saturation and Formation of Clouds . . . . . . . 32
2.6 Dry and Moist Adiabatic Lapse Rates . . . . . . . . . . . . . . . . . . . . . . . . 34
2.7 Stability of Moist Air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3 Observations of Tropical Climate Dynamics and Convectively

Coupled Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1 Historical Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2 Clouds in the Tropics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3 The Madden-Julian Oscillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4 Convectively Coupled Equatorial Waves . . . . . . . . . . . . . . . . . . . . . 50
3.5 Self-similar Multiscale Convective Systems . . . . . . . . . . . . . . . . . . 52
xi
xii Contents

4 Introduction to Stochastic Processes, Markov Chains, and Monte

Carlo Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.1 Computing with Random Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.1.1 Law of Large Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.1.2 Monte Carlo Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.1.3 Inverse Transform Method . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.1.4 Acceptance-Rejection Method . . . . . . . . . . . . . . . . . . . . . . . 60
4.2 Introduction to Markov Chains and Birth-Death Processes . . . . . . 61
4.2.1 Discrete Time Markov Chains . . . . . . . . . . . . . . . . . . . . . . . 61
4.2.2 The Poisson Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.2.3 Continuous Time Markov Chains . . . . . . . . . . . . . . . . . . . . . 64
4.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Part II The Deterministic Multicloud Model

5 Simple Models for Moist Gravity Waves . . . . . . . . . . . . . . . . . . . . . . . . 75

5.1 Wave-CISK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.2 A Simple Adjustment Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.3 Mass-Flux Schemes and WISHE Waves . . . . . . . . . . . . . . . . . . . . . 82
5.4 Stratiform Instability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.4.1 Nonlinear Simulations with the Stratiform Model:
The Beautiful and the Ugly . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

6 The Multicloud Model with Congestus Preconditioning . . . . . . . . . . . 99

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.2 The Model Formulation and Main Closure Assumptions . . . . . . . . 100
6.3 Linear Stability Analysis and the Congestus Preconditioning
Instability Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.4 Nonlinear Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

7 Convectively Coupled Equatorial Waves in the Multicloud

Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.1 Governing Equations and Method of Solution . . . . . . . . . . . . . . . . . 118
7.2 Uniform Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
7.3 Meridional Shear Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.4 Effect of a Vertical Shear Background . . . . . . . . . . . . . . . . . . . . . . . 130

8 Convective Momentum Transport and Upscale Interactions

in the MJO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
8.2 CMT in a Simulated Kelvin Wave: A Thought
Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.3 CMT Parameterization Prototypes . . . . . . . . . . . . . . . . . . . . . . . . . . 140
8.3.1 Some Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
8.3.2 Stochastic Model of Majda and Stechmann . . . . . . . . . . . . . 143
Contents xiii

8.3.3
CMT Parameterization with a Third Baroclinic Feedback
and its Effect on an MJO like disturbance . . . . . . . . . . . . . . 146
8.4 Multiscale Waves in MJO Envelope and CMT Feedback . . . . . . . . 151
8.4.1 A Simple Multiscale Model with Features of CMT . . . . . . 151
8.4.2 Equatorial Waves in a Realistic MJO Background . . . . . . . 155

9 Implementation of the Multicloud Model in an Aquaplanet Global

Climate Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
9.1 Introduction: The Cumulus Parameterization Problem . . . . . . . . . . 163
9.2 The Multicloud Model as a Simplified Cumulus
Parameterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
9.3 Uniform Aquaplanet Simulations: MJO v.s. Convectively Coupled
Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
9.4 Warm Pool Simulations: MJO Initiation and Northward
Propagation, Monsoon Climatology, and Variability . . . . . . . . . . . . 172

Part III The Stochastic Multicloud Model: SMCM

10 Stochastic Birth and Death Models for Clouds . . . . . . . . . . . . . . . . . . . 185

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
10.2 A Birth-Death Model for Convective Inhibition . . . . . . . . . . . . . . . 189
10.2.1 The Microscopic Stochastic Model for CIN:
Ising Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
10.2.2 The Coarse Grained Mesoscopic Stochastic Model:
Birth-Death Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
10.2.3 Gillespie’s Exact Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . 195
10.2.4 Numerical Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
10.3 The Stochastic Multicloud Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
10.3.1 The Stationary Distribution, Cloud Area Fractions, and the
Equilibrium Statistics of the Lattice Model . . . . . . . . . . . . . 201
10.3.2 Coarse Grained Birth-Death Stochastic Model and the
Mean Field Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
10.3.3 The Deterministic Mean Field Equations and Numerical
Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
10.4 Coupling the SMCM to a Cumulus Parameterization . . . . . . . . . . . 209
10.4.1 The SMCM in a Toy-GCM . . . . . . . . . . . . . . . . . . . . . . . . . . 210
10.5 Inference of SMCM Parameters from Data . . . . . . . . . . . . . . . . . . . 215
10.5.1 The Bayesian Inference Procedure . . . . . . . . . . . . . . . . . . . . 217
10.5.2 The Giga-LES Inferred Time Scales . . . . . . . . . . . . . . . . . . 219
10.6 SMCM with Nearest Neighbour Interactions . . . . . . . . . . . . . . . . . . 221
10.6.1 The Multiple Particle Hamiltonian . . . . . . . . . . . . . . . . . . . . 223
10.6.2 Coarse Grained Approximation . . . . . . . . . . . . . . . . . . . . . . 225
10.6.3 Mean Field Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
10.6.4 Numerical Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
xiv Contents

11 Implementation of the SMCM in a Global Climate Model . . . . . . . . 229

11.1 SMCM-HOMME Aquaplanet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
11.1.1 Case of a Uniform Forcing . . . . . . . . . . . . . . . . . . . . . . . . . . 230
11.2 Role of Stratiform Heating on Organized Convection . . . . . . . . . . . 233
11.2.1 MJO and Kelvin Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
11.2.2 Case of Asian Monsoon-Like Warm Pool . . . . . . . . . . . . . . 234

12 SMCM in CFS: Improving the Tropical Modes of Variability . . . . . 245

12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
12.2 Implementation of SMCM in CFS . . . . . . . . . . . . . . . . . . . . . . . . . . 247
12.2.1 Prescribed Vertical Profiles of Moistening/Drying . . . . . . . 251
12.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
12.3.1 Improved Climatology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
12.3.2 Improved Tropical Modes of Variability . . . . . . . . . . . . . . . 254
12.3.3 MJO and Monsoon ISO Propagation . . . . . . . . . . . . . . . . . . 257
12.3.4 Further Physical Aspects of the MJO Signal . . . . . . . . . . . . 263
12.3.5 Rainfall Event Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 265

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Acronyms

LCL Lifted condensation level

LFC Level of free convection
LNB Level of neutral buoyancy
CAPE Convective available potential energy
CAPEl Low-level CAPE
CIN Convective inhibition
GCM General circulation (or Global climate) model
HOMME High-order method modeling environment
CFSv2 Climate Forcast System version 2
WRF Weather Research and Forecasting Model
CRM Cloud resolving model
LES Large eddy simulation
ITCZ Intertropical convergence zone
MCMC Markov chain Monte Carlo
MJO Madden-Julian oscillation
ISO Intra-seasonal oscillation
ABL Atmospheric boundary layer
WIG Westward inertia-gravity wave
EIG Eastward inertia-gravity wave
RCE Radiative-convective equilibrium
CMT Convective momentum transport
MCM Multicloud model
SMCM Stochastic multicloud model
MRG Mixed Rossby-gravity wave
MGW Moist gravity wave
MGWI Moist gravity wave instability
BSISO Boreal summer intra-seasonal oscillation
MISO Monsoon intra-seasonal oscillation
LPS Monsoon low-pressure system
CCEW Convectively coupled equatorial wave
CISK Convective instability of the second kind
xv
xvi Acronyms

WISHE Wind-induced surface heat exchange (also called wind-

evaporation feedback)
WTG Weak temperature gradient approximation
GARP-GATE or GATE Field experiment: Global Atmospheric Research Pro-
gram - GARP Atlantic Tropical Experiment
TOGA-COARE Field experiment: Tropical Ocean Global Atmosphere
Coupled Ocean-Atmosphere Response Experiment
CINDY-DYNAMO Field experiment: Cooperative Indian Ocean Experiment
on Intraseasonal Variability—Dynamics of the MJO
 Part I
Background and Preliminaries
 Part I introduces some basic background material which is important for the
understanding of the subsequent-core chapters of this book. The expert reader can
skip directly to Part II.
Chapter 1
The Governing Equations and Dry
Dynamics

1.1 Introduction

The equations of motion that govern atmospheric (and also oceanic) flows are based
on the theory of fluid mechanics comprising the Euler and/or Navier Stokes equa-
tions which model conservation of mass, momentum, and energy [27, 11] of a
Newtonian fluid such as air and water. The so-known hydrostatic primitive equa-
tions are derived from these basic laws of physics after some major simplifications
or approximations taking into account the particular topology of planetary flows
[61, 216, 166]. Namely, the lower atmospheric layer, known as the troposphere,
where weather and important climatic systems take place, is very thin compared to
the extent of Earth’s surface and as such horizontal flows (parallel to Earth’s sur-
face) are much more important on the global scale than vertical motions. Vertical
motions are much weaker on planetary and synoptic scales (∼1000 km - 5000 km)
and strong vertical flows are more confined to smaller convective scales on the or-
der of the tropospheric depth. Thus, for the flows of interest for global climate and
medium- to long-range weather modelling, for instance, the vertical acceleration is
neglected in the momentum equations resulting in what is known as the hydrostatic
balance equation
ρ 1 ∂ p
g =− , (1.1)
ρ0 ρ0 ∂ z
where g is the gravity acceleration constant, ρ  and p are density and pressure per-
turbations from fixed background values ρ0 and p0 . The hydrostatic approximation
can be derived by simple scaling arguments based on the above observations [48].
This equation should not be confused with the static pressure equation which states
that the vertical pressure gradient, at any given height, acts as a constant reaction
force balancing the weight of the air column above that point. Instead it expresses

© Springer Nature Switzerland AG 2019 3

B. Khouider, Models for Tropical Climate Dynamics, Mathematics of Planet Earth 3,
https://doi.org/10.1007/978-3-030-17775-1 1
4 1 The Governing Equations and Dry Dynamics

the deviations from this background due to large-scale wave motions, for example.
The total air pressure and total air density satisfy,

∂ p0
ρ = ρ0 + ρ  , p = p0 + p , gρ0 = − > 0.
∂z
Note the > inequality sign in this equation which contrasts with (1.1) where the
pressure gradient can have an arbitrary sign. The quantities ρ0 and p0 are often
referred to as a background stratification which is removed from (1.1) for conve-
nience.
Another important simplification, which also results from the small aspect ra-
tio of the tropospheric layer, amounts to neglecting the variations of density in the
momentum and mass conservation equations, except for the buoyancy term. This is
known as the Boussinesq approximation [48, 61, etc.]. Moreover, for the diabatic
flows of concern, density variations are due mostly to temperature variations except
for pressure changes due to adiabatic lifting and sinking. It is thus convenient to in-
troduce the potential temperature, θ , as the temperature a parcel of air would have
if it is displaced adiabatically (without exchange of heat with its environment) to a
reference pressure ps (usually the surface pressure). Using the equation of state for
an ideal gas, p = Rρ T , where T is temperature and R is the gas constant, and the
first law of thermodynamics, namely that the change in internal energy at constant
pressure is balanced by the work done by pressure, Cp dT = ρ −1 d p, we arrive at
[48, 285]  κ
ps
θ= T, (1.2)
p
where κ = RC−1p ≈ 0.286 with C p is the heat capacity at constant pressure of dry air.
If we express the ideal gas law in terms of potential temperature for both the actual
environment and the background state, then we obtain

p ρ θ p
1+ = (1 + )(1 + )(1 + )κ
p0 ρ0 θ0 p0
or
ρ θ p
= − + (1 − κ ) + h.o.t
ρ0 θ0 p0
where h.o.t is the combination of all high order terms of the Taylor expansion, in
terms of ρ  /ρ0 , p /p0 , θ  /θ0 ,which are assumed to be relatively small. If, in addi-
tion, we ignore the term involving p /p0 under the grounds that pressure perturba-
tions adjust quickly due to fast sound waves, then to a first order approximation, the
hydrostatic balance in (1.1) can be rewritten as

θ 1 ∂p
g = . (1.3)
θ0 ρ0 ∂ z
We note that in (1.3) θ0 can be taking to be a constant independent of height, θ0 =
300 K, without aggravating the underlying approximation.
1.1 Introduction 5

θ0
The vertical gradient of the background stratification ddz defines the stability of
the background state for when it is positive it relates to a situation where lighter
(warmer) fluid lies on top of heavier (colder) fluid. In this case we have a stable
stratification. The normalized quantity

1 d θ0
N2 = g
θ0 dz

is the square of the so-called Brunt-Vaı̈saı̈la buoyancy frequency. When N 2 is pos-

itive, it defines the frequency at which a vertically displaced, by an infinitesimal
amount, parcel of air would oscillate around its initial position. Indeed, if z is the
vertical displacement of the parcel, from its original position, zi = 0, then Newton’s
second law states
θ0 (z) − θ0 (zi ) 1 d θ0
z̈ = −g ≈ −zg ⇐⇒ z̈ + N 2 z = 0,
θ0 θ0 dz
which is indeed the equation of motion – harmonic oscillator with the oscillation
frequency N. In the free troposphere, above the neutrally stable mixed boundary
layer (and below the tropopause), N is approximately constant. Unless, otherwise
stated, here, we set N = 0.01 sec−1 , which is equivalent to a potential temperature
increase in height of roughly 3 K km−1 , consistent with observations [61, 48].
Geophysics texts often refer to the lapse rate or dry adiabatic lapse rate (more
on this in Chapter 2), denoted by Γd as being the rate at a which a parcel of (dry)
air cools down as it is raised adiabatically, i.e., along the curve of constant θ . The
value of Γd can thus be derived directly from (1.2) by making use of the hydrostatic
equation gρ = − ddzp . The case of moist air is slightly different and will be discussed
separately in the next chapter. We have
 κ  
dθ ps dT g
0= =κ − =⇒ Γd = C−1 p g ≈ 10 K km .
−1
dz p dz Cp
dT0
In some textbooks the quantity S = dz − Cgp is called the static stability, where T0 is
the actual background or environmental temperature profile. Note that N 2 ∝ S. The
given background is said to be stable if S ≥ 0 and unstable if S < 0. When S = 0
the profile is said to be neutrally stable. The latter situation is often observed in the
mixed planetary boundary layer (PBL).
In addition to the force of gravity, which induces important wave motions known
as gravity waves, which in turn play a central role of this book, another important
force is due to Earth’s rotation and is known as the Coriolis force. When the equa-
tions of motion are written in a frame of reference which rotates with Earth’s angular
velocity Ω = 1 day−1 , the most significant component of the Coriolis force points
in the vertical, and acts on the moving fluid as a force in the direction perpendicular
to its horizontal velocity everywhere but at the equator where this effect is zero, be-
cause the rotation axis is perpendicular to Earth’s radius there. Much like the force
of gravity, which induces hydrostatic balance dynamics, the Coriolis force induces
6 1 The Governing Equations and Dry Dynamics

balanced flows, where this force is balanced by the horizontal pressure gradient, to-
wards which atmospheric and oceanic flows adjust relatively rapidly, especially in
mid- and high-latitudes where this force is significant. Similarly to the hydrostatic
balance, the so-called geostrophic balance relation is behind an important approx-
imation, known as quasi-geostrophic dynamics, leading to quasi-two-dimensional
flows where a single variable, known as the potential vorticity, is conserved and
plays the role of relative vorticity in 2D Euler equations and fully determines the
flow field [61, 166, 216]. Much like the hydrostatic balance induced oscillations,
deviation from geostrophy induces synoptic to planetary scale waves, the scales at
which Coriolis force is important, known by the generic name of Rossby waves,
which are responsible for most of the mid-latitudes weather patterns.
Because the Coriolis force vanishes at the equator, the quasi-geostrophic approx-
imation breaks down in the tropics and leads the way to much richer and more
complex dynamics. As we will see in the next sections and throughout this volume,
because of this property, the equator acts as a wave-guide to a large family of waves
that travel in both directions along the equator. These equatorially trapped waves in-
teract with moist convection and water vapour, nonlinearly with each other, and with
extra-tropical Rossby waves. During the early development of meteorology as a sci-
entific discipline, tropical dynamics and tropical weather were thought to be chaotic
and unpredictable but it is now widely accepted that most of tropical precipitation
is associated with synoptic and planetary scale waves coupled with convection, i.e.,
cloud systems. The two-way coupling between these waves and convection and their
contribution to global dynamics are the main topic of this volume.

1.2 The Primitive Equations

In addition to the hydrostatic (1.3) and the Boussinesq approximation mentioned

above, we introduce the beta-plane approximation for the equations of motion on a
rotating spherical coordinate system [61]; the Earth’s surface is assumed to be flat
along a certain strip surrounding the globe parallel to the equator. The effects of
Earth’s curvature are ignored except for the Coriolis parameter which is assumed to
take the linear form f (y) = β y, where y is the geodesic distance from the equator.
The resulting equations, known as the hydrostatic equatorial beta-plane primitive
equations, then read

∂v ∂v
+ v · ∇v + w + β yv⊥ = −∇p + Sv ,
∂t ∂z
∂ p gθ
= ,
∂z θ0
∂θ ∂θ ∂ θ̄
+ v · ∇θ + w +w = Sθ ,
∂t ∂z ∂z
∂w
∇·v+ = 0. (1.4)
∂z
1.2 The Primitive Equations 7

Here, v = (u, v)T is the horizontal velocity field, where u is the zonal (along the
equator) component, v is the meridional (North-South) component, and w is the
vertical velocity. The corresponding coordinates, x, y, z, which represent longitude,
latitude, and altitude, are directed from West to East, from South to North, and from
bottom to top, respectively, while t > 0 is time. We fix z = 0 at the surface or bottom
of the troposphere and z = H is at the top of the troposphere (H = 16 km). The
operator ∇ = (∂x , ∂y ) is the horizontal gradient and X · Y is the dot product of the
two-dimensional vectors X and Y . Also v⊥ = (−v, u) is the perpendicular velocity
vector—a rotation by 90 degrees to the left.
The scalars p and θ represent the pressure (per unit mass) and potential tempera-
ture perturbations, respectively. The equations in (1.4) have been normalized by the
density, which is assumed constant except when multiplied by the gravity acceler-
ation, according to the Boussinesq approximation. The total potential temperature,
including a background that depends only on height, is given by

θ total (x, y, z,t) = θ0 + θ̄ (z) + θ (x, y, z,t)

where θ0 = 300 Kelvin is a reference constant temperature and θ̄ defines the vertical
profile-background stratification and satisfies N 2 = θg0 ∂∂θ̄z > 0 where N is the Brunt-
Väisälä buoyancy frequency.
The first equation in (1.4) represents the conservation of horizontal momentum
with β = 2Ω /R ≈ 2.2804 × 10−11 m−1 s−1 is the gradient of the Coriolis parame-
ter at the equator. Here, Ω = 2π /24 hr−1 is the frequency of Earth’s rotation and
R = 6378 km is its radius. The second equation expresses hydrostatic balance, and
the third one is the energy equation written in terms of the potential temperature.
The last one is the divergence free constraint which expresses conservation of mass,
often called the continuity equation, under the Boussinesq approximation. The terms
on the right-hand side, Sv and Sθ , are the sources and sinks of momentum and en-
ergy, respectively. In particular, Sθ represents convective heating from clouds and
radiative cooling (and heating), which constitute the main sources of energy input
in the tropics.
The equations in (1.4) are augmented with the rigid lid boundary conditions

w|z=0,H = 0. (1.5)

While this assumption is certainly valid at the surface z = 0, it is only approxi-

mately valid at the top of the tropopause; It is merely justified by the sharp transition
of the static stability there, making the stratosphere much more stable and leaving
the weather flows confined to the tropospheric layer. Gravity waves of various sorts
do penetrate to the stratosphere and stratospheric flows are believed to influence the
climate system on longer time scales through processes such as the quasi-biennial
oscillation and radiative forcing [68, 10]. However, these two-way interactions be-
tween the troposphere and the stratosphere can be ignored for our purpose. Periodic
boundary conditions are naturally assumed in the x-direction with a period equalling
the equatorial circumference of the globe while they are left somewhat arbitrary in
the y-direction; either vanishing solutions at y −→ ±∞ or no-flow boundaries at
8 1 The Governing Equations and Dry Dynamics

y = ±L for some distance L from the equator will be considered when dealing with
practical situations, case by case.

1.3 Equatorially Trapped Waves

In this section we will derive some special solutions for the linearized version of
the primitive equations (1.4), on the form of plane waves along the equator. They
are known as the equatorially trapped waves because as they travel zonally, along
the equator, they are meridionally confined to the vicinity of the equator over some
distance set by a dimensional parameter known as the Rossby deformation radius.
Before we make a formal mode expansion in the vertical direction in order to devise
a systematic Galerkin truncation methodology expanding the equations (1.4) into
a set of vertical mode equations, we will first use a method developed in [187] of
transforming the primitive equations into a set of shallow water equations. Notice
that the intuition behind this transformation comes from the fact that the primitive
equations are meant to model planetary flows with horizontal scales of hundreds of
kilometres while their vertical scales are limited by the tropospheric height of up to
16 km. Indeed, from the perspective of these planetary motions, the troposphere is
a thin layer of fluid surrounding the solid Earth.

1.3.1 The Shallow Water Approximation

We consider the state of rest, u = v = w = 0, with a constant background stratifica-

tion, set by the buoyancy frequency N = 0.01 s−1 [61], as our equilibrium solution
about which the equations of motion are linearized and simple wave solutions are
sought. With the nonlinear advection terms, v · ∇v, w ∂∂ vz , v · ∇θ , w ∂∂θz , and the forc-
ing terms, Sv , Sθ , all set to zero, we combine the three last equations to obtain an
evolution equation for the pressure.
  
∂ ∂ 1 ∂p ∂w
+ = 0. (1.6)
∂ t ∂ z N 2 (z) ∂ z ∂z

From dimensional analysis point of view, the differential operator ∂∂z N12 ∂∂z acts
on the pressure perturbation p and transforms it into a dimensionless field. It is thus
reasonable to look for special solution for which this operator is merely a measure-
ment of the pressure perturbation in terms of a reference scale given by gH ∗ , which
is simply the hydrostatic pressure per unit mass of a hypothetic atmospheric column
of height H ∗ . In other words, H ∗ sets the scale of the pressure perturbation under
consideration. Mathematically, we need to choose H ∗ so that λ = gH1 ∗ is an eigen-
value of our differential operator and the special solutions are simply the associated
eigenfunctions:
1.3 Equatorially Trapped Waves 9
 
∂ 1 ∂p ∂p
− = λ p, 0 ≤ z ≤ H; = 0, z = 0, H. (1.7)
∂z N ∂z
2 ∂z

We will postpone the choice of H ∗ and the computation of these eigenfunctions

for a moment and we will first look at the implications of this simplification on their
horizontal variations.
For a fixed scale height H ∗ , the unforced-linearized primitive equations reduce
to the shallow water-like equations,

∂u ∂p
−βy v = −
∂t ∂x
∂v ∂p
+βy u = − (1.8)
∂t ∂y
 
∂p ∂ u ∂ v
+ c2 + = 0,
∂t ∂x ∂y

where for convenience the u and v equations were written separately and the last
equation is nothing but (1.6) combined with the continuity
√ equation. For a reason
to be clarified later, we also set c2 = gH ∗ , where c = gH ∗ is known as the gravity
wave speed, according to the standard theory of shallow water equations. We note
however that one of the main differences between (1.8) and the standard shallow
water equations resides in the physical meaning of the constant H ∗ . While in the
standard case, the equivalent of H ∗ is the mean water column height, around which
surface waves oscillate and propagate, here H ∗ is typically internal to the typical
tropospheric height, up to H ≈ 16 km, and only sets the vertical scale of the pressure
disturbances that may not be directly seen on top of the fluid column. The associated
waves are called internal waves as opposed to surface waves that are seen at the
beach, for example.
We now return to the eigenvalue problem (1.7) to express the value of the equiv-
alent depth H ∗ and the phase speed c. The general solution is
mπ z mπ √ 2
p(z) = cos( ), = N λ , m = 1, 2, · · · .
H H
This yields λ = π 2 m2 /(H 2 N 2 ) = (gH∗)−1 = c−2 . Thus, we have for m = 1, for
example, H ∗ = 259 m and c ≈ 50 m s−1 . The integer values of n define the various
internal modes also known as the baroclinic modes. More on this in the next section,
where we consider a more systematic mathematical treatment and thus justification
of the elliptic eigenvalue problem.

1.3.1.1 The Kelvin Wave

We begin by looking at the special case when the meridional velocity is zero, v = 0.
In particular, this reduces the system in (1.8) to a single equation for u alone, since p
and u become constrained by the v-equation. As we will see throughout this volume,
despite its simplicity the resulting solution plays a major role in tropical dynamics—
10 1 The Governing Equations and Dry Dynamics

it is perhaps the most important of all tropical waves observed in nature. The v-
equation yields a balance between the meridional pressure gradient and the zonal
velocity,
∂p
βy u = − .
∂y
The remaining equations for u and p suggest a travelling wave solution along the
equator. Through separation of zonal and meridional variables x and y, we let

u(x, y,t) = f (x,t)φ (y), v = 0, and p(x, y,t) = g(x,t)ψ (y).

Substituting this ansatz in (1.8) yields,

ft φ (y) = −gx ψ (y),

β y f φ (y) = −gψy ,
gt ψ (y) + c2 fx φ (y) = 0.

The first and third equations suggest that φ and ψ are constant multiples of each
other and without loss of generality, we can assume ψ = φ . Thus, combining the
two equations leads to the wave equation,

ftt = c2 fxx ,

which has simple solutions consisting of both left going and right going waves of
the form f (x,t) = h(x ± ct) and g(x,t) = ∓c h(x ± ct). The second equation then
gives an equation for φ .
βy
φy = ± φ ,
c
whose elementary solution takes the form, φ (y) = exp ± β2cy . The plus sign corre-
2

sponds to a westward propagating wave is unbounded as |y| −→ ∞ while the minus

sign yields the solution
 
1 β y2
u(x, y,t) = p(x, y,t) = h(x − ct) exp − . (1.9)
c 2c

The associated flow perturbation structure is displayed in Figure 1.1 when the
2π
function h takes a wave form h(z) = sin(kz), for a fixed wavenumber k = 8000
−1
km , or an 8000 km wavelength. Here, the gravity wave speed is set to c = 50
m s−1 , corresponding to the first baroclinic mode. As expected, the main charac-
teristics of the Kelvin wave include a perfect symmetry about the equator, eastery
winds (winds pointing to the West) are aligned with a low-pressure disturbance and
westerlies (winds pointing to the East) are on top of a high pressure ridge. And more
importantly, the wave is confined to the vicinity of the Equator (y = 0) with both the
winds and pressure anomalies being strongly attenuated as we move away from the
tropical region. The wave is said to be trapped in the vicinity of the equator. The dis-
tance by which the wave gets attenuated is determined by the dimensional parameter
1.3 Equatorially Trapped Waves 11

Le = c/β ≈ 1480 km1 , when c = 50 m s−1 or an equivalent depth H ∗ ≈ 255 m. Le

is known as the equatorial Rossby deformation radius. It sets a natural length scale
for synoptic scale dynamics which are responsible for weather disturbances in the
tropics occurring on time scales of two to ten days.

1.3.1.2 Rossby, Gravity, and Mixed Rossby-Gravity Waves

We now return to the general case with v = 0, and introduce the reference scales c =
√ ∗ −1
gH for velocity, Le = c/β for length, and Te = Le /c = cβ for time. When
written in these units the equation in (1.8) will remain unchanged except for the
constants β and c2 which will factor out leaving the same set of equations but with
β and c2 being both unity. In some sense, the chosen reference scales are the scales
at which the beta-effect and gravity wave dynamics have equal importance. It has
to be pointed out however that different equivalent depths can result in substantially
different gravity wave speeds and consequently different length and time scales.

Fig. 1.1 Pressure anomalies (contours, negative anomalies are dashed) and velocity profile (ar-
rows) for the Kelvin wave in (1.9). Here, c = 50 m s−1 , h(z) = sin(kz), k = 2π /8000 km−1 .

1 Le = 2c/β , in some textbooks.

12 1 The Governing Equations and Dry Dynamics

As noted above, the value H ∗ ≈ 255 m (c = 50 m s−1 ) yields Le ≈ 1480 km and

consequently Te = 8.2263 hours. However, for H ∗ = 64 m we get c = 25 ms−1 ,
Le = 1047 km, and Te = 11.6338 hours. Smaller equivalent depth leads to waves
with slower propagation speeds, smaller wavelengths, and longer time scales. As
can be expected this will have important implications on the interaction dynamics
of waves of different equivalent depths.
We seek simple wave solutions on the form u(x, y,t) = û(y) exp [i(kx − ω t)],
v(x, y,t) = v̂(y) exp [i(kx − ω t)], and p(x, y,t) = p̂(y) exp [i(kx − ω t)] with i2 = −1.
When this ansatz is inserted into (1.8), we get the following eigenvalue problem:

−iω û − β yv̂ + ik p̂ = 0
−iω v̂ + β yû + p̂ = 0 (1.10)

−iω p̂ + c (v̂ + ikû) = 0.
2

The primed variables denote y-derivatives. The three equations in (1.10) are com-
bined to obtain a standalone equation for v̂.
 2 
 ω β 2 y2 kβ
v̂ + 2 − k − 2 −
2
v̂ = 0,
c c ω
ω û = iβ yv̂ + k p̂, (1.11)

(ω − c k ) p̂ = i c2 kβ yv̂ − ω c2 v̂ .
2 2 2


By the change of unknowns v̂(y) = cH( βc y) exp − β2cy , the first equation re-
2

duces to
H  − 2ξ H  + μ H = 0,

for H(ξ ) with μ + 1 = c

β
ω2
c2
− k2 − kωβ
= Te2 ω 2 − Le2 k2 − c ωk or simply, μ + 1 =

ω 2 − k2 − ωk in the chosen non-dimensional frame of reference, and ξ = βc y is
the non-dimensional distance from the equator. This is the well-known Hermite’s
equation which admits an infinite number of polynomial solutions provided μ is an
even non-negative integer [144]. This results in an infinite set of solutions,
     
β β y2 c ω2 kβ
v̂n = cHn y exp − , − k 2
− = 2n + 1, n = 0, 1, 2, · · ·
c 2c β c2 ω
(1.12)

The Hermite polynomials are given by

dn
Hn (ξ ) = (−1)n exp(ξ 2 ) exp(−ξ 2 )
dξ n

and satisfy the recursive relations

Hn+1 (ξ ) = 2ξ Hn (ξ ) − Hn (ξ ), n = 0, 1, 2, · · · ,
1.3 Equatorially Trapped Waves 13

Hn (ξ ) = 2nHn−1 (ξ ), n = 1, 2, · · · (1.13)

The first few are given by

H0 (ξ ) = 1, H1 (ξ ) = 2ξ , H2 (ξ ) = 4ξ 2 − 2, · · ·

The restriction to polynomial solutions for Hermite’s equation guarantees that v̂

vanishes exponentially fast, as e−ξ /2 , in the same fashion as the Kelvin wave solu-
2

tion in (1.9). Nonetheless, the Hermite polynomials modulated by exp(−ξ 2 /2) (the
functions defining the v̂n solutions in (1.12)) constitute a complete set for square
integrable functions, so in principle any other physically sound solution can be ex-
pressed as a (possibly infinite) sum of such solutions. The modulated Hermite poly-
nomials are called the parabolic cylinder functions and will be discussed in some
more details below.
The relationship between frequency and wavenumber in (1.12) is called the dis-
persion relation and it can be used to express, for instance the phase speed, cs = ω /k,
and the group velocity, cg = ddkω , of the waves as functions of the wavenumber. While
the phase speed determines the propagation speed of peaks and troughs of individual
waves, the group velocity merely determines the speed of wave packets or rather the
energy carried by the waves [61]. When the phase speed varies with k so that waves
of different wavelengths travel at different phase speeds, the associated waves are
said to be dispersive. For dispersive waves, the group velocity and the phase speed
are distinct. The integral parameter n in (1.12) fixes the mode of meridional structure
and the shape of the dispersion curves. It is often called the meridional index.
For fixed values of n and k, ω is determined by solving the cubic equation in
(1.12). There are three distinct solutions. For n ≥ 1, they are approximately given
by2

k
ωr ≈ − , ωge = k2 + 2n + 1, ωgw = − k2 + 2n + 1. (1.14)
k2 + 2n + 1
The first solution, ωr , corresponds to slowly (low-frequency) westward-moving
waves known as equatorial Rossby waves by analogy to the geostrophic Rossby
waves found in midlatidudes. The other two correspond, respectively, to eastward
and westward moving inertio-gravity waves, they are the equivalents of Poincaré
waves [61].
The case n = 0 is somewhat special. The corresponding dispersion relation
√ easily
factors out to yield the three solutions: ω0 = −k and ω± = 12 k + ± 12 k2 + 4. The
first solution yields a non-dispersive westward moving, Kelvin-like wave, which is
discarded [216]. In fact, if this solution is inserted back into (1.11), then it yields
v ≡ 0 which takes us back to the case of the Kelvin wave considered above. The
two other solutions corresponding to a westward moving and an eastward moving
waves, depending on the sign of k, behave like Rossby and eastward inertio-gravity
waves, respectively. They are often called Yanai waves [294, 229, 253] but they also
2 Consider the two cases when ω << 1 and when ω >> 1 separately to simplify the dispersion
relation to approximately −k2 − ωk = 2n + 1 and ω 2 − k2 = 2n + 1, respectively.
14 1 The Governing Equations and Dry Dynamics

bear the name mixed Rossby-gravity waves because they branch out accordingly
depending on the sign of k.
The dispersion relation for the Kelvin wave can be recovered from (1.12) by
setting n = −1 to yield the single solution ω = k. To help visualize the different
cases, the dispersion relation curves corresponding to the first few values of n (n =
−1, 0, 1, 2, 3) are plotted in Figure 1.2.
Once the meridional velocity v is fixed by the parabolic cylinder function corre-
sponding to a given meridional mode n ≥ 0, the pressure and zonal velocity fields
can be recovered through the last two equations in (1.11). The flow structure of the
Yanai and the n = 1 gravity and Rossby waves are displayed in Figure 1.3. They
should be compared with that of the Kelvin wave in Figure 1.1. It is worthwhile

Fig. 1.2 Dispersion relation curves of equatorially trapped waves corresponding to the first few
meridional indices, n = −1, 0, 1, 2, 3, comprising the Kelvin, the Yanai (or mixed Rossby-gravity),
Rossby, and inertio-gravity waves. Here, the equivalent depth H ∗ = 255 m and the frequency units
are in cycles per day (CPD) and k is expressed in terms of the planetary wavenumber corresponding
to a maximum wavelength P = 40000 km (k = 1), the Earth’s perimeter at the Equator.
Another random document with
no related content on Scribd:
 The streets of old Cairo resound with the cries of vendors of sweetmeats and
drinks. Lemonade is dispensed from a great brass bottle on the back of the seller,
while around his waist is a tin tray of glasses or cups.
 Over many warehouses, shops, and even stables of old Cairo are homes of the
well-to-do with marble floors covered with fine rugs. The supporting arch is much
used because long timbers are not available.

Here is a lady with a eunuch, who, as black as your hat and as

sombre as the Sphinx, guards the high-born dame lest she should
flirt with that handsome young man from Tunis sitting cross-legged in
the midst of his bottles of attar of roses. He offers a bottle to the lady
while he talks of its merits in the most flowery terms. Here is a
barefooted girl, who, strange to say, has no veil over her face, but
whose comely features might be considered by a jealous lover to
warrant such protection. Her chin is tattooed and the nails of her
fingers and toes are stained deep orange with henna. She has a
great tray on her head and is calling out her wares in the strangest
language: “Buy my oranges! They are sweet as honey, and I know
that God will make my basket light.”
This is in Arabic, and one hears the same extravagant sort of talk
all about him. Here two Turks meet and salute each other. They
almost fight in their struggle each to humble himself first by kissing
the hand of the other. After they have done so a third passes and
they all say: “Naharak sayed”—“May thy day be happy and blessed.”
There are no more polite people on earth than these
Mohammedans, whose everyday talk is poetry.
I can always amuse myself for days in watching the trading in the
bazaars. I saw an Egyptian woman buying some meat to-day. The
butcher’s whole stock consisted of a couple of sheep, one of which
hung from a nail on the wall. The woman drew her finger nail along
the piece she wished to take home, and the butcher sawed it off with
a clasp knife. He weighed it on a pair of rude scales, and the woman
objected, saying that he had given her too much. He then took one
end of the strip of meat in his hand, and putting the other end in his
mouth, severed it by drawing the knife quickly across it. He handed
the piece he had held in his mouth to the woman, who took it and
paid for it, evidently seeing nothing out of the way in his methods.
In the bazaars the merchants sit in little booths no bigger than the
packing-box of a piano. A ledge about two feet high, and of about the
same width, runs along the front of the store, on which the
customers sit. A purchaser is usually offered coffee, and asked to
take a smoke out of the long-stemmed water pipe of the proprietor. It
takes a great time to make a deal, for the Mohammedan always asks
three times what he expects to get, and never comes down without
bargaining. The better merchants all keep book accounts, which they
foot up in Arabic characters, taking the ink out of a brass inkstand
with a handle a foot long which is so made that it will contain the pen
as well as the ink. This inkwell is thrust into the belt of the gown
when the proprietor leaves his shop.
If one is not satisfied at one place he can go to another. In the
Cinnamon Bazaar there are dozens of stores that sell nothing but
spices, and in the Shoemakers’ Bazaar are the gorgeously
embroidered slippers and red-leather shoes, turned up at the toes,
worn by all good Mohammedans. In the Silver Bazaar the jewellers
are at work. They use no tools of modern invention. Their bellows is
a bag of goatskin with a piece of gun-barrel for the mouth and two
sticks like those used for the ordinary fire bellows at the end. One’s
only guarantee of getting a good article is to buy the silver, have it
tested by the government assayer, and let the jeweller make it up
under his own eyes. Poor jewellery is often sold, and I remember
buying a silver bracelet for a friend during a visit to Cairo which
looked very pretty and very barbaric, but six months after its
presentation it began to change colour, and proved to be brass
washed with silver.
I see many watches displayed, for there is now a craze among the
peasants of Egypt to own watches. They want a cheap article, and in
many cases buy a fresh watch every year. As a result the Swiss and
Germans have been flooding the country with poor movements, put
up in fancy German silver, nickel, and gun-metal cases, and are
selling them at two dollars and upward apiece. They are not equal to
our timepieces which sell at one dollar. Some of these watches are
advertised as of American make, and sell the quicker on that
account. I doubt not that a good American watch would sell well and
displace the poor stuff now sent in by the Swiss. In one bazaar only
brass articles are shown, while in another nothing but rugs are sold.
The Persian Bazaar and the Turkish Bazaar are managed by men of
these nations. In fact, wandering through the business parts of Cairo,
one can see types of every oriental people on the globe.
 CHAPTER VIII
INTIMATE TALKS WITH TWO KHEDIVES

To-day Egypt is governed by a king. Her last sovereign had the

title of sultan, and for fifty years before that she was ruled by
khedives. There were four khedives in that time, and with two of
them I had face-to-face chats. The first was with Tewfik Pasha,
whom I met in the Abdin Palace during my second visit to Cairo. The
other was with Abbas Hilmi, the son and successor of Tewfik, with
whom I talked sixteen years later. Abbas Hilmi’s pro-German
intrigues finally led to his being deposed by the British and to the
establishment of the Protectorate, which ended in the nationalization
of Egypt under a ruler with the title of king.
I give you here the stories of the two interviews, reproducing the
notes I made at the time.

I have just returned from a long audience with the Khedive of

Egypt. Khedive is a Persian-Arabic word, meaning “king,” and
Mohammed Tewfik occupies much the same position now as the
Pharaohs did in the days of Moses. It is true that he is in a measure
the vassal of the Sultan of Turkey to whom he pays a tribute of about
three and three quarter million dollars a year, and that he has also
several European advisers who keep sharp watch over the revenues
of his kingdom to see that a great part of them go to the interest on
the debts that he and his predecessors have contracted with the
bankers of Europe. But he is, nevertheless, the king of Egypt, and as
kings go to-day, he has more power than many other monarchs. His
residence in Cairo is a grand palace with hundreds of rooms filled
with magnificent furniture. He drives about the city with soldiers
carrying swords, riding prancing horses in front of his carriage, and
with a score of cavalry following behind. He has five hundred
thousand dollars a year for his personal expenses, and he has
several palaces besides the one he occupies in Cairo.
It was at the Abdin Palace that I met His Highness to-day. The
interview had been arranged by the American consul general. We
left his office together in the consular carriage. The dragoman of the
legation, a bright-eyed Syrian in the most gorgeous of Turkish
clothes of brown covered with gold embroidery and with a great
sword shaped like a scimitar clanking at his side, opened the
carriage door for us and took his seat by the coachman. The Arabian
Jehu cracked his whip and away we went through the narrow
streets. We drove by the modern European mansions of the rich
Greeks, past the palaces of Egyptian princes from which came the
sweet smell of orange flowers and over which whispered broad
spreading palms. We then went through a business street amid
droves of donkeys, through a caravan of camels, by veiled women
clad in black, past the palace in which Ismail Pasha had his harem
when he was khedive, and on into a great square of many acres. On
the right of this square were vast barracks filled with Arab troops in
blue uniforms and fezzes. A regiment of Egyptian troops was going
through a gymnastic drill, performing the motions as well to-day as
they did at the time when our American General Stone was their
commander and when General Grant reviewed them and said that
they seemed to be good soldiers for everything except fighting.
The Abdin Palace, built in the form of a great horseshoe, is at the
end of this square. It is a vast building of two stories, of brown
stucco, with many windows and a grand entrance way in the centre.
At the left there is a door leading to the harem, and as our carriage
drove up we were passed by a closed coach drawn by two
magnificent Arabian horses. On the box beside the liveried
coachman sat a scowling eunuch whose black skin and dark clothes
were all the more sombre by contrast with his bright red skullcap. In
front of the carriage ran two fleet syces with wands or staffs held up
in the air in front of them, warning plebeians to get out of the way. I
was told that the carriage was that of a princess who was about to
make a call upon the Khedivieh, or queen. These runners, who are a
part of every nobleman’s turnout, are among the most picturesque
sights of Egypt.
At the door of the palace stood two pompous soldiers with great
swords in their hands. They were in Turkish costumes with
embroidered jackets of blue and gold and full zouave trousers of
blue broadcloth. Upon their heads were turbans, and their faces
made me think of the fierce troops that conquered this land in the
days of the Prophet. Passing up the massive steps we came to the
palace door which was opened by an Arab clad in European clothes
and wearing the red fez, which the Egyptian never takes off in the
house or out of it. We were ushered into a grand entrance hall,
floored with marble mosaic, the walls of which were finished in
cream and gold. In front of us a staircase so wide that two
wagonloads of hay could be drawn up it without touching led by easy
flights to the second floor, while at the right and the left were the
reception rooms for visitors and halls leading to the apartments
reserved for the chamberlains, masters of ceremonies, and other
officers of the royal household. After chatting a moment with one or
two of the cabinet ministers, who were just passing out after a
council with His Highness, we moved on up the stairs. In one of the
drawing rooms on the second floor we were met by another Egyptian
official in black clothes and red fez who conducted us to a reception
room, the door of which stood open, and motioned us to enter.
In the centre of this room, which was not larger than a good-sized
American parlour, there stood all alone a man of about thirty-six
years of age. He was dressed in a black broadcloth coat buttoning
close up at the neck like that of a preacher. Lavender pantaloons
showed below this, fitting well down over a pair of gaiter-like pumps.
On the top of his rather handsome head was a fez of dark red with a
black silk tassel. This man was the Khedive of Egypt. He is, I judge,
about five feet six inches in height and while rather thick-set, does
not weigh more than one hundred and fifty pounds. His frame is well
rounded, his head is large, and his features are clean cut. He has a
nose slightly inclined to the Roman. His forehead is high, and the
dark brown eyes that shine from under it change from the grave to
the smiling during his conversation. The Khedive extended his hand
and said he was glad to see me and that he liked to have Americans
come to Cairo. Seating himself on a divan, with one leg doubled up
under him, he motioned me to join him. There was an absence of
pomp or snobbishness in his manner, and though dignified he did not
put on half the airs of the average backwoods member of our House
of Representatives. As he seated himself, his black coat opened so
that I had a chance to note the contrast between his costume and
that of the gorgeous rajahs whom I have met in India. His only
jewellery consisted of a set of pearl studs the size of the smallest of
peas and a watch chain of thin links of gold. He wore a cheap black
bow tie in his white turnover collar, and his cuffs, though
scrupulously clean, had not the polish of the American laundry.
Besides being a good French scholar, Tewfik Pasha speaks
English, and that was the language used in our conversation. In
speaking of his life as Khedive, he said:
“I am told that many people envy me my position. They say that I
am a young man whose lot must be a pleasant one. They do not
understand the troubles that surround me. Many a time I would have
been glad to lay down all the honours I have for rest and peace. The
ten years of my reign have been equal to forty years of work and of
worry. If life were a matter of pleasure I would be a fool to remain on
the throne. I believe, however, that God put man on the world for a
purpose. Duty, not pleasure, is the chief end of man. I do the best I
can for my country and my people, and I feel happiest when I do the
most work and when my work is the hardest.”
 “In the famous Abdin Palace I interviewed Tewfik Pasha, when he was Khedive
of Egypt, and later, in the same audience room, talked with Abbas Hilmi, his son
and successor.”
 The gorgeous kavass is essential to the official dignity of the representative of
foreign governments in Cairo. Besides attending on the person of minister or
consul general on state occasions, he also serves as major domo and general
“fixer.”

As the Khedive said these words I thought of the thorns which

have filled the pillow of his reign. I thought of how, upon his entering
manhood, his father Ismail was deposed and he was put upon the
throne. I thought of how he boxed the ears of the messenger who
came to tell him he had succeeded to that uncomfortable seat. I
thought of his trouble under foreign dictation. I thought of the plots
and nearly successful rebellion of Arabi Pasha, of the revolution of
the Mahdi, of the creditors who to-day are grinding Egypt between
their upper and nether millstones, of the danger of assassination,
and of the other perils that are ever present about the throne of an
oriental monarch. Recalling all these things, I could appreciate why
his mouth hardened and his eyes grew sad when he spoke thus to
me.
The talk then turned upon the condition of Egypt and its future, but
as to these matters Tewfik was reticent. He spoke proudly of the
reforms which he had inaugurated in government and of the fact that
now, though the taxes were heavy, every peasant knew just what he
would have to pay and that the taxes were honestly collected. He
spoke of the improvement of the courts and said that the pasha and
the fellah were equal before the law. “When I came to the throne,”
said he, “the people were surprised that I put the prince on the same
footing as other people. Now, there is no difference in justice. The
prince and the peasant are the same in our courts, and the former
may be punished like the latter.”
At this point, coffee and cigarettes were brought in by the servants
of the palace. The coffee was à la Turque. It was served in little china
cups shaped like egg cups, in holders of gold filigree, each holding
about three tablespoonfuls of rich black coffee as thick as chocolate
and as sweet as molasses. There were neither saucers nor spoons.
Trying to follow the Khedive’s example I gulped down half the
contents of the cup at a swallow. It was as hot as liquid fire. I could
feel the top of my mouth rising in a blister, the tears came into my
eyes, and my stomach felt as though it had taken an internal Turkish
bath. Tewfik Pasha took the boiling mixture without winking and went
on talking as though his throat were used to scalding fluids.
Surprised to see him refuse a cigarette, I asked him if he did not
smoke. He replied:
 “No! I neither smoke nor drink. I do not drink for two reasons. I
believe a man is better off without it, and, what is of more moment to
me, it is against the laws of life as laid down in the Koran. We do not
believe it right to drink anything intoxicating and good Moslems drink
neither wine nor liquor. I believe that every man should be faithful to
the religion which he professes. My faith is that of Islam and I try to
follow it as well as I can. I am not illiberal in it, however, for I tolerate
all religions and all sects in my kingdom. We have Copts, Jews, and
Christians, and your missionaries are at work in the land. They make
very few converts, if any, among the people of my faith, but they
have schools in Upper Egypt that are doing much in the way of
education.”
The consul general here spoke of the Khedive’s knowledge of the
Koran, mentioning the fact that His Majesty knows the whole book by
heart. There is no doubt that Tewfik has as much faith in his religion
as we have in ours. He spoke with some pride of the Mohammedan
conversions in Africa and the fact that there are more than one
hundred millions of people in the world who believe the same as he
does. We talked of the band of one hundred American Catholics,
who are stopping in Egypt on their way to the Holy Land, and the
Khedive said he was interested in these pilgrims who are following
the footsteps of Joseph and Mary. He spoke of the immense sums
brought into Egypt by tourists and said that it bettered the business
of his country.
Throughout our whole conversation the talk was of the most
cordial and unceremonious character and I left the palace with the
impression that the Khedive of Egypt is a man of great sense and of
more than ordinary ability. He stands well with his people. Indeed,
the leading men in Cairo tell me he would do much for Egypt if he
were not hampered by foreign intervention. He gave up a number of
his palaces a year or so ago and he is, for a king, most economical.
Had other rulers of the past been equally careful, Egypt would be a
rich country to-day instead of being ridden with debts. He is a man of
domestic tastes, and though a Mohammedan and an oriental king,
he is the husband of but one wife to whom he is as true as the most
chaste American. A friend of Tewfik Pasha reported to me a talk he
recently had with him upon this subject in which the Khedive
expressed himself strongly in favour of monogamy: “I saw,” said he,
“in my father’s harem the disadvantages of a plurality of wives and of
having children by different wives, so I decided before I came to
manhood that I would marry but one woman and would be true to
her. I have done so, and I have had no reason to regret it.”
From what I can learn the ruler’s family life is a happy one. He is
much in love with his wife, who is said to be one of the cleverest
women of Egypt. A woman friend of hers, who visits often at the
royal harem, tells me that this queen of Egypt is both beautiful and
accomplished. She keeps up a big establishment separate from that
of the Khedive, and when she sits down to dinner or breakfast it is
not with her husband, but with her own ladies. The Khedive eats with
his officers, according to Mohammedan etiquette, and his
apartments, or the salumlik, are separate from hers. Both she and
her husband have done much to break down the rigidity of
Mohammedan social customs. Tewfik Pasha takes the Khedivieh
with him wherever he goes, though she usually travels in a separate
train or car. She has stuck to the Khedive through the stormiest days
of his reign. During the last war she refused to take refuge on the
English gunboats when invited to do so.
Both the Khedive and the Khedivieh are wrapped up in their four
children. They have two boys and two girls. The boys are Abbas
Hilmi, who will be fifteen years old in July, and Mehemet Ali, who is
two years younger. These boys are now at school in Berlin. They
speak French, English, German, and Arabic, and they are, I am told,
very clever. The girls are rather pretty, cream-complexioned maidens
of eight and ten, who are as much like American girls as they can be
considering their surroundings. They wear European clothes and
may be seen along the sea shore at Alexandria, walking together
and swinging their hats in their hands like other little girls at our
summer resorts. They have European governesses and talk French
quite well.
 In Cairo sixteen years later I found on the throne Abbas Hilmi who
was a boy at school when I had my interview with his father. Again
through the courtesy of our consul general an audience with the
Khedive was arranged for me, and together we went to the palace to
pay our respects. Here is the story of my visit:
In the very room where I met Tewfik Pasha I was received in the
same cordial and informal manner by his son, the present Khedive.
He does not look much like his father. He is a trifle taller and seems
to have more dignity, perhaps because in place of his father’s simple
garb Abbas Hilmi wears the more formal frock coat and striped
trousers of modern officialdom.
 Though stripped of most of their political powers, the khedives surrounded
themselves with all the trappings of rulership, and made the most of the
magnificence of the Abdin Palace in Cairo, where they granted audiences and
gave grand balls.
 One of the most famous hotels in the world is Shepheard’s, at Cairo, through
which for many years leading characters of all nations have passed on their way to
the East or to the West. Its site was once part of the garden of Princess Kiamil,
daughter of Mehemet Ali.

A school among the Moslems is a simple matter, consisting usually of young

men sitting at the feet of a teacher whose sole textbook and equipment are the
Koran, lengthy passages of which are learned by rote.

My conversation with His Highness covered a wide range. It dealt

with the present prosperity of Egypt, and I could see that he
understands both his country and its people. He thinks that the Nile
valley has by no means reached the maximum of its development,
and says that by increasing the dams and drainage facilities Egypt
might yield much greater crops than she does now. I spoke to him
about having met his father, mentioning the great interest that Tewfik
Pasha showed in Egypt and its future. The Khedive expressed a
similar desire to do all he could for the Egyptians, but practically the
only matters in which he has full sway are those regarding his own
estates, his management of which shows great business capacity.
He has an allowance of five hundred thousand dollars a year out of
the public treasury, but in addition he owns thousands of acres of
valuable lands, so his private property must be worth many millions
of dollars. He handles this in such a way that it pays well, his
experiments and improvements being the talk of farmers and
business men throughout the Nile valley.
I have heard a great deal of these khedivial farms since I have
been in Egypt. Abbas Hilmi inherited much land from his father, but
he has other large tracts, which he himself has redeemed from the
desert, and yet others which he has made good by draining. Not far
from Cairo he owns twenty-five hundred acres which a few years
ago were covered with swamps, quagmires, and hillocks. He bought
this cheap and then began to improve it. He cut down the hills,
drained the swamps, and put water on the land. At present that
estate is paying over sixty thousand dollars a year, bringing His
Highness thirty per cent. and upward on his investment.
He has another great farm not far from Alexandria which was all
desert not long ago. The Khedive has irrigated it and thus turned four
thousand waste acres into cultivated fields. Farm villages have
grown up about them and His Highness has so laid out the estate
with trees and flowers that it is said to be like an earthly paradise. In
one place he has a plantation of fifteen thousand mulberry bushes,
the leaves of which furnish food for his silkworms. This estate is at
Montzah, a few miles out of Alexandria, on a beautiful bay of the
Mediterranean Sea. Abbas Hilmi has built a palace there, or rather
two palaces, a little one for himself and a larger one for his family. In
other parts of the estate he is carrying on all sorts of breeding
experiments. He has chicken houses and rabbit hutches as well as a
tower containing thousands of pigeons.
The Khedive is interested in fine stock and is doing much to
improve that of Egypt. On his various farms he has high-bred horses,
cattle, and sheep. He has a large number of Arabian thoroughbred
horses, and some Jersey, Swiss, and other fine breeds of cows. His
water buffaloes, known here as gamoushes, are far better than any
others of the Nile valley. He is also breeding cattle for oxen and
mules for draft animals. He has a school on his estate near Cairo
where two hundred boys are being educated to take places on his
various properties. This school is run at his own expense, the boys
being taught farming and surveying as well as reading, writing, and
arithmetic. The course of study lasts for five years, at the end of
which the graduate is pretty sure of a good position as a steward or
overseer on one of the khedivial farms.
Abbas Hilmi has made a great deal of money within the last three
or four years. He is investing largely in Cairo and is building
apartment houses with elevators, telephones, electric lights,
bathrooms, and all other modern improvements. He has a brick
factory on one of his estates near here, and his profits from cotton
and other crops must be very large.
Abbas Hilmi’s wife is the Princess Ikbal Hanem, whom he married
when he was about twenty. She is said to be both accomplished and
beautiful, but like all Mohammedan ladies, she leads a secluded life,
and does not appear at the great functions at the palace. She is not
seen at the Khedive’s grand ball, given to his officials and the
foreigners about once a year, to which something like fifteen hundred
guests are invited. She is present, all the same, however, for she has
a screened chamber looking down upon the ballroom, with the
curtains so arranged that she can watch the dancing and flirting
while she herself is unseen. Her Majesty has gorgeous apartments
in each of the palaces and a little court of her own of which the noble
ladies of Egypt are a part.
 CHAPTER IX
EL-AZHAR AND ITS TEN THOUSAND MOSLEM
STUDENTS

The biggest university of the Mohammedan world is situated in

Cairo. It has, all told, over ten thousand students, and its professors
number more than four hundred. Its students come from every
country where Mohammedanism flourishes. There are hundreds
here from India, and some from Malaya and Java. There are large
numbers from Morocco, as well as from Algeria, Tunisia, and Tripoli.
There are black Nubians, yellow-skinned Syrians and Turks, and
boys from southeastern Europe with faces as fair as our own. There
are long-gowned, turbaned Persians, fierce-eyed Afghans, and
brown-skinned men from the Sudan and from about Kuka, Bornu,
and Timbuktu. The students are of all ages from fifteen to seventy-
five, and some have spent their lives in the college.
This university has been in existence for almost a thousand years.
It was founded A.D. 972, and from that time to this it has been
educating the followers of the Prophet. It is to-day perhaps the
strongest force among these people in Egypt. Ninety-two per cent. of
the inhabitants of the Nile valley are Mohammedans and most of the
native officials have been educated here. There are at least thirty
thousand men in the public service among its graduates, while the
judges of the villages, the teachers in the mosque schools, and the
imams, or priests, who serve throughout Egypt are connected with it.
They hold the university in such high regard that an order from its
professors would be as much respected as one from the
government, if not more.
 A fifteen-minutes drive from the hotel quarter through the bazaars of the Mouski
and the narrow “Street of the Booksellers” brings one to the university of El-Azhar,
for 900 years the educational centre of the Moslem world.