=Paper=
{{Paper
|id=Vol-2534/62_short_paper
|storemode=property
|title=Time-Space Earth Surface Gradients for Satellite Monitoring of Ecosystems
|pdfUrl=https://ceur-ws.org/Vol-2534/62_short_paper.pdf
|volume=Vol-2534
|authors=Alexei V. Kartushinsky,Nikita A. Ogorodov,Alexander A. Larko
}}
==Time-Space Earth Surface Gradients for Satellite Monitoring of Ecosystems==
https://ceur-ws.org/Vol-2534/62_short_paper.pdf
Time-Space Earth Surface Gradients
for Satellite Monitoring of Ecosystems
Alexei V. Kartushinsky(1,2), Nikita A. Ogorodov(3,2), Alexander A. Larko(1)
(1)
Institute of Biophysics SB RAS, Krasnoyarsk, Russia, kartushka@mail.ru
(2)
Siberian Federal University, Krasnoyarsk, Russia, lantar@inbox.ru
(3)
Federal Research Center KSC SB RAS, Krasnoyarsk Russia, Nogorodvv@yandex.ru
Abstract. The results of Earth surface gradients characteristics calculate based on satellite
data are presents. Dynamics structural features of horizontal gradient fields in aquatic
objects and land ecosystems by physical, biological parameters are considered. Problems of
the parametrization for environment exchanges evaluations with using numerical modeling
based on satellite data and software tools are submitted. Spatial-temporal scaling and
averaging of gradient components are discussed. Develop improved estimates of Earth
surface gradients fields in the different ecosystems is considered.
Keywords: satellite data, regional features, gradient fields, frontal zones, ecotone, spatial
averaging, average seasonal variability, synergistic effect.
1 Introduction
Gradient indicators of measurable surface parameters of terrestrial and aquatic ecosystems show some spatial
boundaries of various dynamic processes. Calculations of natural objects gradient characteristics are necessary to
identify zones with different dynamic activity [1, 2]. The satellite measurements in visible and infrared ranges are
images with spatial fields of various natural objects on the Earth's surface. The main parameters in our case are surface
reflectivity and radiation temperature. A methodological concept is determined by a need to calculate spatial gradients
of Earth's surface measurable characteristics. The calculated gradients with different spatial and temporal averaging are
necessary to identify zones with different dynamic activity of physical and biological fields, land vegetation, geological
indicators such as terrain or soil cover.
The gradient characteristics of parameters that are used to study surface dynamics of ecosystem spatial
heterogeneities are differentiated indicators of values change rate. A gradient value in spatial coordinates provides an
opportunity to obtain a vector of development or attenuation of a process. Gradient indicators of spatial variability of a
studied object (field) are often not fully used, especially for quasistatic objects such as land ecosystems. In this case,
calculations of studied fields dynamics are based on a combination or composition of source images (pictures)
compared in absolute quantitative indicators, averaged and interpolated values [3, 4].
Determining development conditions and scale of hydrological processes is necessary for monitoring an ecological
state of aquatic systems including oceans, seas, lakes and waterbodies. Calculation of spatial gradients from satellite
data allows us to assess spatial and temporal scales of formation of zones with high gradient values. The possible
conditions of acting factors and physical mechanisms of hydrobiological and hydrological components interaction can
be studied using methods of numerical simulation based on calculation of partial derivatives of heat transfer, mass, and
pollution diffusion equations with using satellite and contact data [5, 6, 7].
An analysis of Earth's surface state parameters time-space gradients variability is necessary to assess a dynamic
process in aquatic systems and on the solid Earth's surface. Therefore, the main goal of this work is to implement a
method of calculating Earth's surface gradient characteristics from satellite data using developed information
technology and software with the application of mathematical statistics methods.
2 Materials and methods
Satellite data obtained by various scanners such as SeaWIFS, CZCS, AVHRR, MODIS, SPOT, LANDSAT,
AQUARIUS were used to calculate gradient characteristics in various regions of the Global Ocean and Eurasia. The
main attention is paid to identifying patterns of horizontal highly gradient zones formation according to physical and
biological components of the Earth's surface. To do this, use a gradient method of processing satellite data using
appropriate information technology and software [1]. A methodology to study variability of surface gradient fields of
aquatic and terrestrial ecosystems as dynamic objects based on calculation of spatial gradients (latitudinal, meridional
and absolute) with different averaging periods (week, month, season, year) is used. Source data in the work are satellite
images and data of spectroradiometers obtained by high- and low-resolution hardware such as AVHRR MCSST, CZCS,
Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 Interna-
tional (CC BY 4.0).
SeaWIFS, MIRAS AQUARIUS, MODIS, MSS for various time periods from NOAA, TERRA, AQUA, SPOT-4,
LANDSAT-8 satellites.
ENVI software product was used to calculate gradient fields of land vegetation, and IDL programming language
was used to write calculation logic.
3 Summary and discussion
Calculation of spatial gradients of Earth's surface characteristics in latitudinal and meridional directions allows us to
identify areas with significant differences which can be interpreted as boundary or anomalous.
3.1 Gradient fields in the surface Ocean
Gradient fields of hydrological parameters in aquatic systems allow us to estimate intensity of dynamic processes
variability. An interaction of different water masses with different properties forms hydrological fronts that are
determined exactly by calculations of relevant gradients (Figure 1). In this case, the Global Ocean is a dynamic system
in which there are hydrophysical and hydrobiological structures formed by processes of various spatial and temporal
scales.
(A) (B)
Figure 1. А structure of temperature frontal zone of interaction system between the cold Labrador Current and the
warm current Gulf Stream according to calculation of surface ocean temperature absolute (modulo) gradients for 2000;
A - winter; B - summer.
Currently, considerable attention is paid to processing of observational data and their re-analysis using general ocean
circulation models to improve global models of interaction in an atmosphere-ocean-land system [9, 10, 11]. It is
important to understand a trigger for formation of such phenomena as North Atlantic Oscillations (NAO), El Niño-
South Oscillations (ENSO). An analysis of ocean high-gradient zones dynamic activity, determination of ocean’s
temperature field local interaction effects and formation of atmospheric circulations (cyclones, anticyclones) relevant
response based on gradient indicators will make it possible to understand time-space scales of local processes that
determine synergism of global (planetary) phenomena.
The use of spatial gradient indicators calculated from satellite data in mathematical models gives a magnitude of
change rate in a surface layer structure. An analysis of dimensions of heat transfer, mass, and impurity diffusion
equations terms, based on theory of similarity, makes it possible to introduce initial, boundary conditions and a grid
calculation scale in mathematical models when conducting numerical simulations. Numerical modeling based on a 2D
model of heat transfer allows us to study variability of an ocean temperature frontal zones structure, depending on
dynamic conditions of heat transfer, mass and impurity concentration.
3.2 Gradient fields in the surface Land
As for land surface and terrestrial ecosystems, it is important to determine spatial scales of ecotones (transitional
boundaries between ecosystems) and time scales of variability of such zones. Processes with a low rate of dynamic
activity in space such as soil erosion, vegetation cover, and Earth's surface morphology are practically not evaluated by
gradient fields [1]. However, these processes turn out to be significantly related in long-term series of observations.
So, when evaluating biomass of gross primary production (GPP) and net primary production (NPP) of terrestrial
ecosystems using satellite data, it is sometimes difficult to determine the area of individual ecosystems and their spatial
boundaries (ecotones). In addition, in many works, spatial distribution of dynamics of NPP or another indicator is not
considered, but trends averaged over area are estimated [8]. The obtained average results during further verification by
their terrestrial data are often very ambiguous. Therefore, we can talk about inapplicability of linear models for analysis
of time series in complex, spatially distributed systems. Therefore, there is a question of verifying and analyzing
applicability of various methods for obtaining time trends, a detailed examination of their features, strengths and
weaknesses, as applied to Earth remote sensing data included in an NPP assessment model [12].
Calculation of gradient fields of land vegetation indicators, in particular, the NDVI gradients, can help determine
spatial boundaries of ecosystems that in many respects correspond to terrain. The speed and direction of changes in soil
and vegetation cover may have different coherence, which is a response to climate change.
An example of a gradient field of a vegetation index calculated based on a use of space data GIMMS (Global
Inventory Modeling and Mapping Studies) NDVI from 1982 to 2006, half-month composite images, 8-km resolution
(on the equator line), global scale, NOAA satellite AVHRR scanner, presented for the Baikal region in Figure 2.
Figure 2. NDVI gradient field calculated modulo (absolute gradients).
Thus, as a result of applying the gradient approach, it became possible to isolate large changes in boundary
transition between biomes and ecosystems such as taiga – tundra transition or zones of high mountain regions and
lowlands with corresponding types of vegetation.
3 Conclusion
Calculations of gradient characteristics of ecosystems surface based on satellite data make it possible to identify
zones with different dynamic activity. An analysis of gradient fields value distribution makes it possible to identify
heterogeneous and homogeneous zones of ecological systems and, with an appropriate averaging period, to obtain
degree of dynamism of such zones. This allows us to study natural systems structural organization processes and time-
space scales of variability that can be used in development of statistical and deterministic mathematical forecast models.
For aquatic systems, studying a structure of gradient fields of water temperature, salinity, chlorophyll concentration,
turbidity makes it possible to identify a scale of zones of non-uniform characteristics distribution, to determine lifetime
of boundary (transition) zones by magnitude of gradient in meridional, latitudinal direction and modulo. Such a
methodology, combined with a use of appropriate information and software tools, makes it possible to evaluate effect of
factors of a physical and biological nature on ecosystem formation as a whole or to assess cyclical nature of these
factors action in individual parts of a system. For land ecosystems, degree of surface heterogeneities variability is much
lower than in aquatic systems, however, the gradient approach can provide a differentiated picture of ecosystem
boundaries based on satellite data. Therefore, it is possible to evaluate a synergetic component in self-organization of a
systemic level of interaction between ecosystems and elements of the Biosphere system at an Atmosphere-Ocean-Land
level with relevant scales of time-space averaging.
References
[1] Kartushinsky A.V., Kukoba N.A. Investigation of gradient fields of the earth surface based on satellite data //
Vestnik SibGAU. 2015. Vol. 16. P.587-596.
[2] Kartushinsky A.V., Sidorenko A.Y. Analysis of the variability of temperature gradient in the ocean frontal zones
based on satellite data // Advanced Space Research. 2013. Vol.52, No. 8. P. 1467–1475.
[3] Kobernichenko V.G., Trenikhin V.A. Methods of image synthesis based on Earth remote sensing data of various
resolutions // Successes of Modern Radio electronics. 2007. № 4. P. 22-31.
[4] Planeta Monitoring. Programmnyj kompleks obrabotki sputnikovyh dannyh / V.V. Asmus, Buchnev A.A.,
Krovotyntsev V.A. and others // Computer Science Issues. 2013. №3. P. 85-99. (In Russian)
[5] Kartushinsky A.V. The investigation on the dynamics of frontal zones in the ocean based on the numerical model-
ing, using the AVHRR satellite data // Advances in Space Research. 2004. 33 (7). P. 1173-1178.
[6] Kartushinsky A.V. Dynamic Features of Frontal Zones Structure in the Ocean for Using in the Numerical Models
Based on Satellite Data // Journal of Siberian Federal University. Mathematics & Physics. 2011. 4(2). P. 208–216.
[7] Kartushinsky A.V. Chislennoe modelirovanie effektov gidrofizicheskogo vozdejstviya na raspredelenie fitoplank-
tona / Matematicheskaya biologiya i bioinformatika. 2012.Vol. 7. №1. P. 112-124. (In Russian)
[8] Larko A.A., Ivanova Yu.D., Shevyrnogov A.P. Nelinejnye trendy chistoj pervichnoj produkcii rastitel'nosti yuga
Krasnoyarskogo kraya // Fundamental'nye issledovaniya. 2015. №3. P. 106-110. (In Russian)
[9] Kazmin A.S. Izmenchivost krupnomasshtabnyh okeanicheskih frontalnyh zon: analiz globalnoj sputnikovoj
informacii // Sovremennye problemy distancionnogo zondirovaniya Zemli iz kosmosa. 2012. Vol. 9. № 1. P. 213-
218. (In Russian)
[10] Serykh I.V. O dinamike i strukture globalnoj atmosfernoj oscillyacii v klimaticheskih modelyah i realnosti //
Okeanologicheskie issledovaniya. 2018. Vol. 46. P. 14-28. (In Russian)
[11] Rayner N.A., Parker D.E., Horton E.B. Global analyses of sea surface temperature, sea ice, and night marine air
temperature since the late nineteenth century //JGR: Atmospheres. 2003. Vol. 108, No. 14. 4407.
[12] Ivanova Yu. D., Bartsev S.I., Pochecutov A.A., Kartushinsky A.V. The analysis of seasonal activity of photosyn-
thesis and efficiency of various vegetative communities on a basis NDVI for modeling of biosphere processes /
Advances in Space Research, 39 (1), 2007, pp. 95-99.