Bulgarian Academy of Sciences. Space Research and Technology Institute.

Aerospace Research in Bulgaria. 35, 2023, Sofia

DROUGHT MONITORING USING REMOTE SENSING DATA

Antoaneta Frantzova
Geological Institute – Bulgarian Academy of Sciences
e-mail: afrantzova@gmail.com

Keywords: Vegetation indexes, Remote sensing, Comparative analysis, Drought

Abstract
The paper deals with the application of the remote sensing for drought monitoring. A way of
their application in practice are shown through example for monitoring of drought that is done for
three different area in Bulgaria with different pattern, for the same time (July) and for eleven randomly
selected years from 1994 to 2009. For the purpose, NDVI, LAI and FPAR vegetation indexes are used..
According to Intergovernmental Panel on Climate Change (IPCC) drought and floods are the most
severity events threatening the country so attention is paid on drought monitoring.

Introduction
Theoretical Description of Vegetation Indices
The theoretical basis for ‘empirical-based’ vegetation indices is derived
from examination of typical spectral reflectance signatures of leaves. The reflected
energy in the visible is very low as a result of high absorption by photosynthetically
active pigments with maximum sensitivity in the blue (470 nm) and red (670 nm)
wavelengths. Nearly all of the near-infrared radiation is scattered (reflected and
transmitted) with very little absorption, in a manner dependent upon the structural
properties of a canopy (LAI, leaf angle distribution, leaf morphology). As a result,
the contrast between red and near-infrared responses is a sensitive measure of
vegetation amount, with maximum red - NIR differences occurring over a full
canopy and minimal contrast over targets with little or no vegetation. For low and
medium amounts of vegetation, the contrast is a result of both red and NIR changes,
while at higher amounts of vegetation, only the NIR contributes to increasing
contrasts as the red band becomes saturated due to chlorophyll absorption.
The spectral reflectance of vegetation is detectable in three major EMS
regions [1, 2]:
1. Visible region (400–700 nm) – Low reflectance, high absorption, and
minimum transmittance. The fundamental control of energy-matter interactions with
vegetation in this part of the spectrum is plant pigmentation.

52
 2. NIR (700–1350 nm) – High reflectance and transmittance, very low
absorption. The physical control is internal leaf structures.
3. MIR (1350–2500 nm) – As wavelength increases, both reflectance and
transmittance generally decrease from medium to low, while absorption increases
from low to high.

Method
Drought monitoring using NDVI NOAA AVHRR data
NOAA AVHRR and MODIS (Terra/Aqua satellites) data are used for
drought monitoring. They can be summarized in: MOD13 A1 (250 m 16 days NDVI,
250 m 16 days EVI); MOD13 A2 ( 1 km 16 days NDVI, 1 km 16 days EVI)
MOD15A1 (Fpar 1 km, Lai 1 km); MOD17 (Npp 1km); MOD44A (Land Cover
Change Metrics Past 3 Months)
Land cover monitoring using NOAA NDVI is done by calculations of NDVI
for three different regions with various land patterns for following years: 1994, 1997,
2000, 2002, 2005, 2007, 2009, 2010 2005, 2007 and 2009: region 1 – Central North
Bulgaria (fig. 5), region 2- Central Bulgaria, around Chirpan city (fig.6) and part of
Rodopi mountain (Eastern Rodopy) – region 3 (fig. 2). Selected regions are shown
on the image below.
Normalized Difference Vegetation Index provides information of vegetation
health and a means of monitoring changes in vegetation over time. NDVI is
calculated from the visible and near-infrared light reflected by vegetation. Healthy
vegetation absorbs most of the visible light that hits it, and reflects a large portion of
the near-infrared light. Unhealthy or sparse vegetation reflects more visible light and
less near-infrared light. Calculations of NDVI for a given pixel always result in a
number that ranges from minus one (-1) to plus one (+1); however, no green leaves
gives a value close to zero. A zero means no vegetation and close to +1 (0.8–0.9)
indicates the highest possible density of green leaves.

Fig. 1. Vegetation anomaly in 2007(earthobservatory.nasa.gov) and selected areas

53
 NOAA AVHRR NDVI database used for land cover monitoring contains
monthly and weekly reference images for the period from April to October (weekly
image is composed by 21 images – 3 images per one day). Database is made with
the data published in DLR EOWEB and the data received in former Space
Monitoring Center at the Ministry of Interior (Bulgaria). Because of the large volume
of data for the selected regions, in the report are presented only the result for Rodopy
mountain region (Region 3). The pixel values are presented in DN from 0 to 255.
Each DN value corresponds to NDVI value. Calculations of NOAA NDVI for a
given pixel always result in a number that ranges from minus one
to 0,7.

Fig. 2. DMC satellite images. Region 3, July 2008 and 2011. Each selected area is divided
on the grid with dimension of 10x10 km.; each one pixel has resolution of 1000 m.
41°37'26.28'', 25°46'13.81''; 41°37'30.57'', 25°53'14.81''
41°32'10.23'', 25°46'19.80''; 41°32'14.50'', 25°53'20.23''

Table 1. DN and NDVI value

NDVI 0–255 NDVI 0–255 NDVI 0–255
0.1 – 0.2 94 0.3 – 0.4 127 – 158 0.5 – 0.6 191 - 218
0.2 – 0.3 95 – 126 0.4 – 0.5 159 – 190 0.6 – 0.7 219 – 254

240

210

180

150
25
1
5
9
13
17
21

29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
97

1994-07 2000-07

Fig. 3. Years with minimal values of NOAA NDVI

54
 270

240

210

180

33

69
1
5
9
13
17
21
25
29

37
41
45
49
53
57
61
65

73
77
81
85
89
93
97
1997-07 2002-07 2005-07 2007-07 2009-07

Fig. 4. Years with maximal values of NOAA NDVI

The average annual temperatures are progressively increased in the recent

decades. They are highest in 1994 - by 1–2 °C higher than average. The maximum
temperature in July 2000 was a record - 40–45 ° C, with 2–4 ° C above the climatic
norms. Minimum temperatures are consistently abnormal, with excesses to 22 °C.
2002 remains the third warmest year in recent years, following by 2000 (according
to the NIMH – BAS).
According to the results, we can conclude that the data from 1994, 2000 and
2007 can be used as lower threshold for monitoring and risk assessment of drought.
For the Eastern Ropodi (Region 3) years with minimal values for July are 1994,
2000, 2002 and 2010 (fig. 3), and the years for maximal values are 1997, 2002, 2005,
2007 and 2009 (Fig. 4).

Table 2. Amounts of rainfall for summer time from 2000 to 2010 (http://eea.government.bg)
Precipitati Precipitatio Precipitation Precipitation Precipitation
on-mm n-mm -mm -mm -mm
Month/Year 2000 2001 2002 2003 2004
June 82.35 93.2 98.6 118 204
July 11.82 118.41 348 78 106
August 8.47 58.712 129 66 128
September 35.07 46.887 165 73 72

Precipitat Precipitat Precipitat Precipitat Precipitat Precipitat

ion ion ion ion ion ion
Month/Year 2005 2006 2007 2008 2009 2010
June 56 219 15 154 85 210
July 178 86 2.8 12.6 55.0 117
August 1 9 96.1 19.5 18.5 7
September 39 57 125.3 29.4 119.6 54

55
 MODIS NDVI, LAI and FPAR comparative analysis and correlation
The aim of the study is to show the relationship between vegetation indices
derived from satellite data, and the way in which they are change for different types
of surfaces. There is a strong relationship between NDVI-LAI, FPAR- NDVI and
LAI-FPAR as is shown on the figures below.
The MOD 15 Leaf Area Index (LAI) and Fraction of Photosynthetically
Active Radiation absorbed by vegetation (FPAR) are 1km at-launch products
provided on a daily and 8-day basis. LAI defines an important structural property of
a plant canopy, namely the one-sided leaf area per unit ground area. FPAR measures
the proportion of available radiation in the photosynthetically active wavelengths
(400 to 700 nm) that a canopy absorbs. The LAI product will be a LAI value between
0 and 8 of the global gridded database. The FPAR product will be an FPAR value
between 0.0 and 1.0 assigned to each 1-km cell of the global gridded data-base. Leaf
Area Index (LAI) is defined as the one sided green leaf area per unit ground area in
broadleaf canopies, or as the projected needleleaf area per unit ground area in needle
canopies. The Fraction of Absorbed Photosynthetically Active Radiation (FAPAR,
sometimes also noted fAPAR or fPAR) is the fraction of the incoming solar radiation
in the Photosynthetically Active Radiation spectral region that is absorbed by a
photosynthetic organism, typically describing the light absoprtion across an
integrated plant canopy. This biophysical variable is directly related to the primary
productivity of photosynthesis and some models use it to estimate the assimilation
of carbon dioxide in vegetation [5]
The LAI and FPAR data are generated by MOD15A2 algorithm, which is a
Level 4 product and is generated automatically every 8 days with a spatial resolution
of 1 km in Sinusoidal projection.
Once generated, the values obtained are stored digitally with a scale-factor
and offset applied to transform the values to their biophysical correspondence. The
equation used to decode the digital values to their analysis form is the following
[7, 8]:
Analyticalpixel = scale factor * (digitalpixel value – offset)
Research and Applications: LAI and FPAR are biophysical variables that
describe canopy structure and are related to functional process rates of energy and
mass exchange. Both LAI and FPAR have been used extensively as satellite derived
parameters for calculation of surface photosynthesis, evapotranspiration, and NPP.
These products are essential in calculating terrestrial energy, carbon, water-cycle
processes, and biogeochemistry of vegetation. The LAI product is an input to Biome
BGC (Biogeochemical) models to produce conversion-efficiency coefficients, which
are combined with the FPAR product to produce daily terrestrial PSN
(photosynthesis) and annual NPP [3, 4].

56
 MOD15A2 provides global LAI and FPAR data derived from the
atmospherically corrected BRDF (MOD 09) using up to 7 spectral ranges
(0.47 µm, 0.555 µm, 0.688 µm, 0.588 µm, 1.24 µm, 2,130 µm).

Fig. 5. Region 2 – Chirpan sity area, Central Bulgaria

42°12'49,86"N, 25°21'9,08"E; 42°12'55,74"N, 25°28'13,90"E
42°7'33,89"N, 25°21'17,27"E; 42°7'39,76"N, 25°28'21,51"E

Fig. 6. Region1 - Central North Bulgaria, DMC image, 32 m, 2008

43°20'49,67"N, 24°52'4,91"E; 43°20'57,51"N, 24°59'17,44"E
43°15'33,86"N, 24°52'15,94"E; 43°15'41,67"N,24°59'27,85"E

57
 The following charts (Fig.7, 8,9) show the relationship between the NDVI,
LAI and FPAR. Theoretically, the relationship between NDVI/FPAR is linear, as
FPAR is inherently derived from NDVI - electromagnetic radiation used to
determine of NDVI, is used to determine of FPAR (photoactive radiation).

90
8900
70
6900 50
4900 30
2900 10
900 -10
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96
7/12/2010/MODIS-LAI-Родопи
7/12/2010/MODIS-LAI-Region 3 7/12/2010/MODIS-NDVI-Родопи
7/12/2010/MODIS-NDVI-Region 3

Fig. 7. Relationship between pixel values for NDVI and LAI for Region 3, July 2010.
(Actual pixel values for NDVI: The values depicted on y-axis have to be divided
into 10,000)

1 100

0,9 90

0,8 80

0,7 70

0,6 60
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96

7/12/2010/MODIS-FPAR-Родопи
7/12/2010/MODIS-FPAR-Region 3 7/12/2010/MODIS-NDVI-Region 3

Fig. 8. Relationship between pixels values for NDVI и FPAR. Region 3, July 2010

58
 120

100

80

60

40

20

0
1
5
9

61

89
13
17
21
25
29
33
37
41
45
49
53
57

65
69
73
77
81
85

93
97
7/12/2010-MODIS-LAI-Родопи
7/12/2011/MODIS-LAI-Region 3 7/12/2010-MODIS-FPAR-Родопи
7/12/2011/MODIS-FPAR-Reg.3
7/12/2010/MODIS-LAI-Region 3
7/12/2011-MODIS-LAI-Родопи 7/12/2011-MODIS-FPAR-Родопи
7/12/2010/MODIS-FPAR-Reg.3

Fig. 9. Relationship between pixels values for LAI and FPAR. Region 3, July 2010
and 2011

It should be noted, in the areas lacking vegetation cover, NDVI and FPAR
have very similar values. In areas with dense vegetation cover with maximum LAI
values, FPAR also reach maximum values. It is obvious that the denser and "green"
vegetation, result in the higher values of the vegetation indices. In our case, logically,
the highest values for all indices are observed for the Rhodope mountain region.

100

80

60

40

20
29

49

69

89
13
17
21
25

33
37
41
45

53
57
61
65

73
77
81
85

93
97
9
1
5

7/10/2010-MODIS-FPAR-Север
7/10/2010/MODIS-FPAR-Region 1

7/10/2010/MODIS-FPAR-Region 2
7/10/201/-MODIS-FPAR-Чирпан
7/10/2010/MODIS-FPAR-Region 3
7/10/2010-MODIS-FPAR-Родопи
Fig. 10. Relationship between pixels values for FPAR indexes, July 2010

59
 Applicability of remote sensing data in the risk management process of
drought

Before
3 3
2 During
After

Fig. 10. Applicability of remote sensing data in the risk management process of drought;
1 – low; 2 – medium; 3 – high

Table 3. Applicability of remote sensing data in the risk management process of drought –
land cover: “before” means – early warning, preparedness, risk and vulnerability
assessment, (including modeling); “during” – monitoring and fast response; “after” –
damage assessment, (including modeling); 1 – low; 2 – medium; 3 – high

Satellite Instrument Before During After Resolution

High resolution Camera system 1 1 1 2–10 m
Middle Camera system
Radometer 3 3 3 10–60 m
resolution
AQUA, TERRA MODIS 3 3 3 250–1100 m
NOAA/POES AVHRR/3 1 1 1 1100 m

Table 4. Applicability of remote sensing data in the risk management process of drought –
water resources
Satellite Instrument Before During After Resolution
High resolution Camera system 1 2 2 2–10 m
Middle Camera system
3 3 3 10–60 m
resolution Radometer
AQUA, TERRA MODIS 1 2 2 250–1000 m
NOAA/POES AVHRR/3 1 1 1 1100 m

Criteria selection for classification are divided in two broad groups. The first
group covers the physical parameters related to particular natural hazards and risk
process in the context of their study and investigation. On the other hand - the second
group is connected with technical characteristics and operating parameter of the
space platforms and technical equipment one the board.

60
 Conclusion
The impact and negative effects of a drought depend on its duration, the
severity and territorial distribution of the deficit, but also to a large extent on the
environment and the socio-economic vulnerability of the affected areas.
Desertification is largely caused by inadequate and unsustainable use of land in
adverse climatic conditions, usually resulting from poverty and lack of livelihoods
and livelihoods.
Land degradation and desertification resulting from prolonged droughts
combined with the socio-economic impact and vulnerability of society are
particularly evident in arid (dry) areas/areas. The global significance of the problem
is underlined by the fact that dry regions cover about 41% of the Earth's surface and
that they are home to about 1/3 of the world's population.
Future climate change scenarios indicate that today's light/medium-sized
drought and drought-related processes are likely to shift to future severe and
prolonged droughts in less than 30 years.
The methodology for land cover monitoring and drought related processes
related to the drought are discussed. It is base on the vegetation indices derived from
satellite data and mathematical algorithms/"products" from satellite data. The main
vegetation indices are considered, and the results of the calculations and comparative
analyzes for several different cases are presented, including different vegetation
indices showing the condition of the vegetation roof and having a direct relation to
its condition and the processes of drought and drought.

References
1. Lusch, D.P. (1999). Introduction to environmental remote sensing. Center for remote
sensing and GIS, Michigan State University, BSRSI, 247 p.
2. Govender, M., Chetty K., Bulcock (2007). A review of hyperspectral remote sensing and
its application in vegetation and water resource studies, Water Research
Commission, AJOL, Vol. 33 No. 2, DOI: 10.4314/wsa.v33i2.49049
3. Global pattern of NPP to GPP ratio derived from MODIS data: effects of ecosystem type,
geographical location and climate
4. Jiang. H. (1999). Modelling the spatial pattern of net primary productivity in Chinese
forests, Ecological Modelling, Volume 122, Issue 3, 20 Oct. 1999, pp. 275–288.
https://doi.org/10.1016/S0304-3800(99)00142-8
5. Christensen, O.B., Goodess, C.M., Harris, I., Watkiss, P. (2011). European and Global
Climate Change Projections. Discussion of Climate Change Model Outputs,
Scenarios and Uncertainty in the EC RTD ClimateCost Project. Summary of Results
from the ClimateCost. Project, funded by the European Community’s Seventh
Framework Programme. Copyright: ClimateCost, 2011, First published September
2011
6. Zhang, H., Xu, M., Chen, H., Adams J. (2009). Global Ecology and Biogeography, (Global
Ecol. Biogeogr.); https://doi.org/10.1111/j.1466-8238.2008.00442.x (2009) 18,
280–290https://www.gnu.org/software/gsl/

61
7. Tucker, C.J., (1979). Red and photographic infrared linear combination for monitoring
vegetation, Remote Sens. Environ., 8, pp. 127–150.
8. Руменина Е. (2008). Научен отчет по договор ДВ 08/001.12.06.2008 г. Разработване
на методология и създаване на национална база данни да NDVI, NPP и LAI на
основата на спътникови данни от NOAA и MODIS. ИКИ, БАН.

ИЗПОЛЗВАНЕ НА ДИСТАНЦИННИ ДАННИ ЗА МОНИТОРИНГ

НА ПРОЦЕСИТЕ НА СУША И ЗАСУШАВАНЕ

А. Францова

Резюме
В настоящата статия се разглежда приложението на дистанционнте
изследвания за наблюдение на процесите, свързани със суша и засушаване.
Начинът на тяхното приложение в практиката е показан чрез пример за мони-
торинг, извършен за три различни района в България, с различни географски
характеристика и за еднин и същи времеви период – м. юли - за период от
единайсет произволно избрани години от 1994 до 2009 г. Използват се
вегетационни индекси NDVI, LAI и FPAR. Според Междуправителствения
панел по изменение на климата (IPCC) сушата и наводненията са най-тежките
събития, застрашаващи територията на страната, поради което настоящето
изследване обръща внимание на мониторинга на сушата.

62