This workshop is sponsored in part by the USA National Phenology Network and the NSF USA-NPN Research Coordination Network (Grant #0639794).

A Workshop on Land Surface Phenology

Presented by

Kirsten de Beurs, Ph.D.Department of GeographyVirginia Tech University

&

Geoffrey M. Henebry, Ph.D., C.S.E.Geographic Information Science Center of Excellence

South Dakota State University

Lecture 1: Introduction to Land Surface Phenology

GIScCEGIScCE

Geoffrey M. Henebry, Ph.D., C.S.E.

Professor of Biology & Geography and Senior Research Scientist

Geographic Information Science Center of Excellence (GIScCE)

South Dakota State University

Geoffrey.Henebry@sdstate.edu

http://globalmonitoring.sdstate.edu

Madison LSP Workshop: 08 APR 2008

1. Contexts for Land Surface Phenology (28)

A. What is Phenology? (16)

B. What are Objects of Phenological Interest? (3)

C. What are Methods of Phenological Observation? (5)

D. What are Methods of Phenological Analysis? (4)

2. Remote Sensing Basics (9)

E. The Planetary Macroscope (3)

F. Radiation Basics (3)

G. Seeing “Green” from Space (3)

3. A Driver of Phenology (18)

H. Whose Time? (2)

I. Réaumur & Thermal Times (6)

J. Example from Konza Prairie LTER: Insolation, AGDD, ANPP (9)

4. Concluding Remarks (3)

Outline of Talk

GIScCEGIScCE

What is phenology ?

Instead every plant must possess a certain adjustment to the season, since the season turns out to be more responsible [for sprouting] than anything else. For all are seen to await their own appropriate season, meanwhile not sprouting at all and not be set in motion, trees and woody and herbaceous plants alike; for the fact is plainest in wild plants, where generation is the plant’s own doing and not promoted by man.

De Causis Plantarum, Book 1, 10.5-10.6

Theophrastus on Plant Phenology (372-288 B.C.)

GIScCEGIScCE

GIScCEGIScCE

Phenology may be defined as the science of investigating the influence of meteorological conditions on the recurrence of the annual phenomena of animal and plant life, such as, for example, the dates of appearance of blossoms or insects.

H.C. Gunton, 1938

Source: Gunton, H.C. 1938. Nature Study Above and Below the Surface: A Bridge between Amateur and Professional. London: H.F.& G. Witherby., LTD. 134pp.

GIScCEGIScCE

Phenology may be defined as the science of investigating the influence of meteorological conditions on the recurrence of the annual phenomena of animal and plant life, such as, for example, the dates of appearance of blossoms or insects.

H.C. Gunton, 1938

Source: Gunton, H.C. 1938. Nature Study Above and Below the Surface: A Bridge between Amateur and Professional. London: H.F.& G. Witherby., LTD. 134pp.

GIScCEGIScCE

Phenology, which derived from the Greek word meaning to show or to appear, is the study of periodic biological events in the animal and plant world as influenced by the environment, especially temperature changes driven by weather and climate.

M.D. Schwartz, 2003

Source: Schwartz, M.D. 2003. Introduction. In: (M.D. Schwartz, ed.) Phenology: An Integrative Environmental Science. Boston: Kluwer Academic Publishers. 564pp.

GIScCEGIScCE

Phenology, which derived from the Greek word meaning to show or to appear, is the study of periodic biological events in the animal and plant world as influenced by the environment, especially temperature changes driven by weather and climate.

M.D. Schwartz, 2003

Source: Schwartz, M.D. 2003. Introduction. In: (M.D. Schwartz, ed.) Phenology: An Integrative Environmental Science. Boston: Kluwer Academic Publishers. 564pp.

GIScCEGIScCE

Phenology has been defined as the study of the timing of recurring biological events, the causes of their timing, their relationship to biotic and abiotic forces, and the inter-relations among phases of the same or different species.

J.Y. Ewusie, 1992

Source: Ewusie, J.Y. 1992. Phenology in Tropical Ecology. Accra: Ghana Universities Press. 109 pp.

GIScCEGIScCE

Phenology has been defined as the study of the timing of recurring biological events, the causes of their timing, their relationship to biotic and abiotic forces, and the inter-relations among phases of the same or different species.

J.Y. Ewusie, 1992

Source: Ewusie, J.Y. 1992. Phenology in Tropical Ecology. Accra: Ghana Universities Press. 109 pp.

GIScCEGIScCE

KEY POINTS

Phenology…

Studies temporal patterns of biological events

Focuses on specific organisms

Seeks to understand the influences on or causes of timing of these events due to:

• abiotic forces

• biotic forces

• intraspecific interactions

• interspecific interactions

GIScCEGIScCE

CAVEATS

Phenology is not…

restricted to annual phenomena

restricted to the extratropics

restricted to temperature effects

restricted to vernal emergence, the “onset of spring”

GIScCEGIScCE

Seasonality is a related term, referring to similar non-biological events, such as timing of the fall formation and spring break-up of ice on fresh water lakes.

M.D. Schwartz, 2003

Source: Schwartz, M.D. 2003. Introduction. In: (M.D. Schwartz, ed.) Phenology: An Integrative Environmental Science. Boston: Kluwer Academic Publishers. 564pp.

GIScCEGIScCE

Some Useful Distinctions & Categories

Phenology is distinct from seasonality: biotic vs. abiotic temporal patterns. But seasonality can affect phenology and vice versa.

Land surface phenology is distinct from phenology: observing electomagnetic radiation at coarse spatial resolution results in a mixture of signals that combines biotic and abiotic components.

Some Useful Distinctions & Categories

Land surface phenologies (LSPs) are the seasonal spatio-temporal patterns of the vegetated land surface [as observed by synoptic sensors at spatial resolutions and extents relevant to meteorological processes in the atmospheric boundary layer].

LSPs affect the timing and magnitude of energy and water exchanges between the land surface and the boundary layer.

GIScCEGIScCE

Key surface types for LSPs:

Ever-green

Spring-green

Rain-green

Never-green

GIScCEGIScCE

GIScCEGIScCE

Differential Land Surface

Phenology due to Land Use

MODIS NDVI @ 1km in 2004

R = 08MAY G = 27JUL B = 08MAR

SD

NE

KS

GIScCEGIScCE

Land Surface Phenology: AVHRR WDRVI 2000 R=Jun02-Jun15, G=Jul14-Jul27, B=Mar10-Mar23

GIScCEGIScCE

What are objects of phenological interest?

Question: What is Spring?—Growth in everything—

Flesh and fleece, fur and feather,Grass and greenworld all together

From The May MagnificatBy Gerard Manley Hopkins

GIScCEGIScCE

GIScCEGIScCE

Species Arrivals & Departures (Appearances & Disappearances)

Species Growth & Development

Species Abundance

Temporal Sequence of Species Interactions, e.g., pollination, seed dispersal.

Temporal Sequence of Population Interactions, e.g., reproductive behavior, mating rites.

Unusual Events, e.g., very early or very late events, disturbance and recovery, invasive species, meteorological extremes.

Others?

GIScCEGIScCE

What are methods of phenological observation?

GIScCEGIScCE

Surveys along routes or transects

Periodic observations in permanent plots

Manipulative experiments with phenological response variables

Citizen science (e.g., GLOBE; NatureMapping)

Remote and proximal sensing

Harvesting of past observations from older articles, monographs, gray literature documents, and theses/dissertations and digitizing them for contemporary analyses.

Others?

GIScCEGIScCE

The Significance of the Observer in Phenology

As implied in the roots of the word, appearance is a critical aspect of phenological investigations, but appearance to whom ?

Survey or sampling design is one key to capturing ephemeral and/or spatially variable phenomena.

Another critical key is training of observers. Phenological items may involve visual or auditory cues. Recognition and correct identification of these cues may require substantial training or prior experience.

Error among observers is rarely quantified.

GIScCEGIScCE

Characteristics of a “good” phenological item pt 1

• low labor cost/simple to observe

• sharp/distinct to minimize error among observers

issue of binomial vs. multinomial vs. ordinal scale

• common/abundant

• high degree of accessibility (visibility or audibility)

Source: Leopold, A., and S.E. Jones. 1947. A phenological record for Sauk and Dane Counties, Wisconsin, 1935-1945. Ecological Monographs 17(1):81-122.

GIScCEGIScCE

Characteristics of a “good” phenological item pt 2

• reliability of recurrence

• continuity

• evidence of newness

• locally-determined dynamics

• sufficient prior knowledge to identify the unusual

• what else?

Source: Leopold, A., and S.E. Jones. 1947. A phenological record for Sauk and Dane Counties, Wisconsin, 1935-1945. Ecological Monographs 17(1):81-122.

GIScCEGIScCE

What are methods of phenological analysis?

GIScCEGIScCE

Database Querying

Graphical Analysis

Association Rules

Exploratory Data Analysis/Data Mining

Statistical Models

Physiologically-based Models of Development

Simulation Models

Interpretative Gestalt

Others?

GIScCEGIScCESource: Leopold, A., and S.E. Jones. 1947. A phenological record for Sauk and Dane Counties, Wisconsin, 1935-1945. Ecological Monographs 17(1):81-122.

Phenology, in short, is a “horizontal science” which transects all ordinary biological professions. Whoever sees the land as a whole is likely to have an interest in it.

A. Leopold & S.E. Jones, 1947

Grasslands phenology emerges from the interactivity of multiple influences

as filtered through the specifics of spatial relationships, genetic heritage, and the process of observation.

source: Henebry, G.M. 2003. Grasslands of the North American Great Plains. In: Phenology: An Integrative Environmental Science (M.D. Schwartz, editor). Kluwer, New York. pp. 157-174.

Using the planetary macroscopeUsing the planetary macroscope

• Learning to see again at different scales.Learning to see again at different scales.• Scale describes the resolution (grain) and extent of Scale describes the resolution (grain) and extent of

measurements measurements andand the relationship between the thing the relationship between the thing observed and its measurement.observed and its measurement.– Spatial scalesSpatial scales– Temporal scalesTemporal scales– Spectral scalesSpectral scales– Radiometric scalesRadiometric scales

• Resisting the naïve interpretations that our evolved visual Resisting the naïve interpretations that our evolved visual cognition system provides.cognition system provides.

Just as telescopes are “light Just as telescopes are “light buckets” gatheringbuckets” gathering starlightstarlight and and

integrating it over extended integrating it over extended periods, periods,

the the constellation of earth-observing constellation of earth-observing sensorssensors gathers and integrates the gathers and integrates the sunlightsunlight andand earthlightearthlight arising from arising from

the planetary surface.the planetary surface.

Remote sensing samples just tiny fractions of Remote sensing samples just tiny fractions of reflected solar & emitted terrestrial radiationreflected solar & emitted terrestrial radiation

Passive SensingPassive Sensing

Shortwave:Shortwave:reflectedreflectedsunlightsunlight

Longwave:Longwave:emittedemitted

earthlightearthlight

IlluminationIllumination

TargetTarget

SensorSensor

GIScCEGIScCE

Any matter with a temperature greater than absolute zero emits electromagnetic radiation.

How to determine what are the characteristics of that emission?

Use the convenient fiction of the “blackbody” a perfect absorber is also a perfect emitter.

GIScCEGIScCE

Half-a-Nickel Course in Radiation Laws

(1) Planck’s Law : Upper bound on the intensity of the radiation emitted by a blackbody:

B(T) d = ( 2 h c2 ) / [ 5 ( e hc / k T -1) ] d

where is wavelengthT is temperature in Kh is Planck’s constant 6.626 x 10-34 J sC is the speed of light in a vacuum 2.998 x 108 m s-1

k is Boltzmann’s constant 1.381 x 10-23 J K-1

Resulting units: W m-2 m-1 sr-1 Power per unit area per unit solid angle per unit wavelength

Intensity of emission contributed by the Intensity of emission contributed by the wavelength interval [wavelength interval [, , +d+d] varies by the ] varies by the inverse fifth power. inverse fifth power. Hotter things emit shorter Hotter things emit shorter ..

Half-a-Nickel Course in Radiation Laws

(2) Wien’s Displacement Law: For a given absolute temperature, Planck’s Law has a single peak at a wavelength that is inversely proportional to that temperature:

max x T = A or max = A / T

where is wavelengthT is temperature in KA is a constant 2897 m K

K oC m

6000 5727 0.48 Sun’s temp

373 100 7.8 H2O boiling pt

310 37 9.3 normal body temp

308 34.4 9.4 APR max (04/22/80)

287 13.7 10.1 APR max normal

273 0 10.6 H2O freezing pt

255 -17.8 11.3 APR min (04/07/82)

Source: Wikipedia

Atmospheric opacity across the electromagnetic spectrum

GDEX Plot 1, Station 1

0

5

10

15

20

25

30

35

400 450 500 550 600 650 700 750 800 850

wavelength (nm)

refle

ctan

ce (%

)

BlueBlue GreenGreen RedRed Near infraredNear infrared

Spectra collected in the Nebraska Sandhills, June 2004Spectra collected in the Nebraska Sandhills, June 2004

Normalized Difference Vegetation Index Normalized Difference Vegetation Index (NDVI) exploits the (NDVI) exploits the spectral contrastspectral contrast between between NIRNIR and and redred reflectance reflectance displayed by (most) terrestrial displayed by (most) terrestrial vegetation.vegetation.

NDVI =NDVI = ( (NIRNIR – – redred) / () / (NIRNIR + + redred))

NDVI gives some indication of the density of NDVI gives some indication of the density of absorbers of photosynthetically active absorbers of photosynthetically active radiation (PAR radiation (PAR 400-700 nm) at the 400-700 nm) at the surface.surface.

However, mapping NDVI into leaf area However, mapping NDVI into leaf area index (LAI) or aboveground net primary index (LAI) or aboveground net primary production (ANPP) is fraught with caveats production (ANPP) is fraught with caveats and uncertainties.and uncertainties.

Nevertheless, it provides an important Nevertheless, it provides an important window into the dynamics of the vegetated window into the dynamics of the vegetated land surface.land surface.

GDEX Plot 1, Station 1

0

5

10

15

20

25

30

35

400 450 500 550 600 650 700 750 800 850

wavelength (nm)

refle

ctan

ce (%

)

GIScCEGIScCE

GIScCEGIScCE

Drivers of Phenology

GIScCEGIScCE

Whose time?

If phenology attends to the timing of recurrent biological events, then whose “clock” and “calendar” should track that time?

Calendars are anthropocentric.

Do the plants pay attention to our calendars?

Can we link vegetation dynamics to a bioclimatological calendar?

GIScCEGIScCE

Heat-sum or degree-day methods are simple and quite old, but have a theoretical foundation in the kinetics of biochemical reactions.

Source: Charles-Edwards, D.A., D. Doley, G. Rimmington. 1986. Modelling Plant Growth and Development. Sydney: Academic Press Australia. 235pp.

GIScCEGIScCE

René Antoine Ferchault de Réaumur (February 28, 1683 - October 17, 1757)

In 1731 he became interested in meteorology, and invented the thermometer scale which bears his name: the Réaumur.

Source: Wikipedia

The freezing point of water is 0 degrees Réaumur, the boiling point 80 degrees Réaumur. Hence, a Reaumur degree is 1.25 Celsius degrees or Kelvins.

Réaumur may have chosen the octogesimal division because the number 80 could be halved 4 times and still be an integer (40, 20, 10, 5); the number 100, for instance, could only suffer this process twice (50, 25).

Réaumur “adopted simply the sum of the mean daily temperature of the air as recorded by a thermometer in the shade, and counting from any given physiological epoch to any other epoch.”

Abbe, Cleveland. 1905. A First Report on the Relations Between Climates and Crops. Bulletin 36. Weather Bureau, No. 342. USDA.

Source: van Keulen, H. 1987. Forecasting and estimating effects of weather on yield. In: (K. Wisiol and J.D. Hesketh, eds.) Plant Growth Modeling for Resource Management: Volume 1 Current Models and Methods. Boca Raton, FL: CRC Press. Pp 105-124. GIScCEGIScCE

“The true problem for the climatologist to settle during the present century is not whether the climate has lately changed, but what our present climate is, what its well-defined features are, and how they can be most clearly expressed in numbers.”

[…]

“It will be seen that rational climatology gives no basis for the much-talked-of influence upon the climate of a country produced by the growth or destruction of forests, the building of railroads or telegraphs, and the cultivation of crops over a wide extent of prairie.”

Cleveland Abbe, Is our climate changing?, Forum, 6, pp 687-688, 1889 Source: http://www.bom.gov.au/bmrc/clfor/cfstaff/nnn/nnn_climate_quotes.htm

The American Meteorological Society has an annual award: The Cleveland Abbe Award For Distinguished Service To

Atmospheric Sciences By An Individual

More on Professor Abbe here: http://www.history.noaa.gov/nwsbios/abbe.html GIScCEGIScCE

GIScCEGIScCE

In phenological studies, it is sometimes convenient to summarize recent weather into measures that can relate to plant growth and development.

Accumulated Growing Degree-Days (also known as thermal time) is a simple biometeorological variable that weights the passage of days by the quantity of “growing degrees” – that portion of the diel temperature range that is useful for plant growth, broadly construed.

The calculation of AGDD is straightforward. The average temperature for a given day is calculated as the simple mean of the maximum and minimum temperatures observed that day. From this average, a “base temperature” is subtracted and values summed:

AGDDt = AGDDt-1 + max[(AverageTempt - Base),0]

GIScCEGIScCE

The magnitude of the base depends on the type of organism under consideration. I have found that for perennial grasslands a base of 0 oC is very effective.

Crop models often use a base of 4 oC for cool season crops (e.g., winter wheat) and 10 oC for warm season crops (e.g., maize).

In the northern hemisphere the convention is to start accumulating at the beginning of the year, with each 01JAN resetting the accumulation. (Is that appropriate?)

While this approach is handy for temperate and boreal systems, it is not very useful in tropical or arid systems in which phenological timing is associated with available moisture, internal carbohydrate pools, co-evolved relationships, or other non-thermal triggers.

GIScCEGIScCE

Temperature serves a surrogate for insolation, but it is also very important itself for regulating rates of enzyme systems.

Thus, temperature is a “mixed signal” and its use is further complicated by the fact that as variable it is intensive rather than extensive.

Extensive measurements scale with change of dimension. Intensive measurements don’t.

Examples: (1) the temperature of this room vs the mass of air in the room; (2) your body temperature vs your body weight.

GIScCEGIScCE

Daily Solar Radiation: Konza Prairie

0

500

1000

1500

2000

2500

3000

3500

4000

0 30 60 90 120 150 180 210 240 270 300 330 360 390

day of year

inso

lation (Jo

ule

/cm

2)

2001

2002

2003

2004

2005

2006

GIScCEGIScCE

Daily Solar Radiation: Konza Prairie

2001 = -0.0685x2 + 25.145x + 198.44

R2 = 0.54

2002 = -0.053x2 + 18.37x + 591.95

R2 = 0.47

2003 = -0.0563x2 + 19.814x + 398.36

R2 = 0.48

2004 = -0.054x2 + 19.111x + 369.67

R2 = 0.51

2005 = -0.0576x2 + 20.893x + 198.1

R2 = 0.54

2006 = -0.0549x2 + 19.071x + 472.97

R2 = 0.52

0

500

1000

1500

2000

2500

3000

3500

4000

0 30 60 90 120 150 180 210 240 270 300 330 360 390

day of year

inso

lation (Jo

ule

/cm

2)

GIScCEGIScCE

Daily Solar Radiation: Konza Prairie

0

500

1000

1500

2000

2500

3000

3500

4000

0 1000 2000 3000 4000 5000 6000

accumulated growing degree-day (base 0 oC)

inso

lation (Jo

ule

/cm

2)

2001

2002

2003

2004

2005

2006

GIScCEGIScCE

Daily Solar Radiation: Konza Prairie -- quadratic fits

2001 = -0.0003x2 + 1.4153x + 1121.1

R2 = 0.55

2002 = -0.0003x2 + 1.1989x + 1106.7

R2 = 0.52

2003 = -0.0003x2 + 1.259x + 1011.2

R2 = 0.53

2004 = -0.0003x2 + 1.1426x + 970.76

R2 = 0.53

2005 = -0.0003x2 + 1.2122x + 879.28

R2 = 0.53

2006 = -0.0003x2 + 1.2146x + 936.61

R2 = 0.56

0

500

1000

1500

2000

2500

3000

3500

4000

0 1000 2000 3000 4000 5000 6000

accumulated growing degree-day (base 0 oC)

inso

lation (Jo

ule

/cm

2)

GIScCEGIScCEUsing a base temperature of 0 oC (273.15 K)

Air temperature is strongly linked to insolation at Konza Prairie

0

1000

2000

3000

4000

5000

6000

0 100000 200000 300000 400000 500000 600000 700000

accumulated daily solar radiation (J oule/cm2)

accum

ula

ted

gro

win

g d

eg

ree-d

ays

(b

ase

0 o

C)

2001

2002

2003

2004

2005

2006

GIScCEGIScCEUsing a base temperature of 4 oC (277.15 K)

Air temperature is strongly linked to insolation at Konza Prairie

0

500

1000

1500

2000

2500

3000

3500

4000

4500

0 100000 200000 300000 400000 500000 600000 700000

accumulated daily solar radiation (J oule/cm2)

accum

ula

ted

gro

win

g d

eg

ree-d

ays

(b

ase

4 o

C)

2001

2002

2003

2004

2005

2006

GIScCEGIScCE

Air temperature is strongly linked to insolation at Konza Prairie

0

500

1000

1500

2000

2500

3000

0 100000 200000 300000 400000 500000 600000 700000

accumulated daily solar radiation (J oule/cm2)

accum

ula

ted

gro

win

g d

eg

ree-d

ays

(b

ase

10 o

C)

2001

2002

2003

2004

2005

2006

Using a base temperature of 10 oC (283.15 K)

AB = -7E-05x2 + 0.4571x - 331.2

R2 = 0.85

UB = -5E-05x2 + 0.3002x - 216.44

R2 = 0.81

0

100

200

300

400

500

600

0 400 800 1200 1600 2000 2400 2800 3200 3600 4000 4400

Accumulated Growing Degree-Day (base 0 oC)

AN

PP

(g

/m2)

GIScCEGIScCE

Konza Prairie LTER biweekly ANPP: 1990-1995

Aboveground biomass dynamics are well approximated by a piece of a parabola.

y = -7E-05x2 + 0.4491x - 327.89

1997 R2 = 0.974

y = -7E-05x2 + 0.4664x - 307.77

1994 R2 = 0.993

y = -8E-05x2 + 0.6339x - 521.24

1993 R2 = 0.978

y = -9E-05x2 + 0.5712x - 408.13

1996 R2 = 0.982

y = -4E-05x2 + 0.4047x - 279.56

1990 R2 = 0.973

y = -4E-05x2 + 0.3879x - 310.23

1995 R2 = 0.974

y = -6E-05x2 + 0.4838x - 396.63

1992 R2 = 0.995

y = -8E-05x2 + 0.5728x - 458.72

1991 R2 = 0.974

y = 1E-05x2 - 0.0284x + 98.699

1989 R2 = 0.7811

0

100

200

300

400

500

600

700

800

900

0 400 800 1200 1600 2000 2400 2800 3200 3600 4000 4400

Accumulated Growing Degree-Day (base 0 oC)

AN

PP

(g

/m2)

Quadratic models can fit well retrospectively, but also fail spectacularly, as in 1989.

Konza Prairie LTER biweekly ANPP on 01A: 1989-1997

GIScCEGIScCE

GIScCEGIScCE

Concluding Remarks

Land Surface Phenology describes the aspects of recurrent biotic and abiotic phenomena that are observable by sensors which have a coarse spatial resolution relative to the objects of interest.

LSPs are not traditional organism-centric specific sets of phenophases. As mixtures of signals, LSPs must be interpreted with care.

One simple, but powerful, approach to modeling temperate LSPs is linking images to weather using thermal time.

2AGDDAGDDNDVI

Peak NDVI is derived from the parameter coefficients and

Green-up period (φ) is the amount of GDD (°C) necessary to reach the peak, and it too can be derived from parameter coefficients.

Quadratic Model of Green-Up

AGDD if

AGDD if ..

3

5051AGDDAGDD

NDVI

Nonlinear Spherical Model of Green-Up

The peak (+) gives the maximum NDVI. is the quantity of AGDD required to reach the peak and corresponds to the duration of the observed green-up phase.

φ =

NDVIpeak =

GIScCEGIScCE

Both LSP models fit well in some regions and neither does in others. Applying the better fit to reanalysis grid cells (~2o) across the northern hemisphere, we can map out four ecologically interpretable metrics that can be derived from the LSP model parameters:

(1) the NDVI at the onset of the observing season;

(2) the seasonal peak NDVI;

(3) the quantity of AGDD needed to reach the peak; and

(4) the seasonal dynamic range of NDVI.

Density sliced images of PC1s from 9 years (1985-88 & 1995-99)

GIScCEGIScCE

QUESTIONS?QUESTIONS?