In this way, the unburned areas aremore likely to remain as a constant translation - In this way, the unburned areas aremore likely to remain as a constant Indonesian how to say

In this way, the unburned areas are

In this way, the unburned areas are
more likely to remain as a constant cover type in the
monthly composite data, rather than undergoing the vegetation–
ice/frost–vegetation cycle seen in the individual
S10 products.
4. Algorithm details
As discussed in Section 3, a significant fall in NIR
reflectance between pre- and post-fire image dates is
regarded as our primary method for mapping burned areas
with VGT. SWIR reflectance also decreases following a fire
event, but we note that this often rapidly recovers a few
weeks after burning, so that in some cases, only a small
decrease in SWIR reflectance is observed between monthly
pre- and post-fire SWIR data. This is demonstrated in Fig.
2c and d, where a 36% fall in SWIR reflectance occurs
between the pre- and post-fire dates, compared to a 62% fall
in the NIR reflectance.
Although Fig. 2 shows that a clear and ubiquitous NIR
reflectance decrease accompanies burning, because other
phenomena unrelated to fire also cause a similar NIR
reflectance change we risk significant commission errors
if we rely on this criterion alone to identify burned areas.
Such phenomena include the residual appearance of ice,
snow, or frost in the pre-fire image composites, which has
melted in the post-fire image, residual cloud shadows that
are present in the post-fire image but not the pre-fire image
(i.e., partly shadowed pixels having erroneously passed the
prior shadow-excluding processing) or the senescence of
vegetation and fall-off of leaves from deciduous trees in
the post-fire image. Further errors can occur due to the
wide swath of the VGT sensor, which can lead to abnormally
high reflectance observations at large viewing zenith
angles due to surface BRDF effects (Stroppiana et al.,
2002). This may cause problems if the pre- and post-fire
data of a pixel were obtained at very different viewing
geometries.
Although the phenomena outlined above may cause NIR
decreases similar to that resulting from fire, further spectral
analysis can help in discriminating ‘true’ burnt areas from
these other effects. A training data set consisting of a
minimum of 15 ‘true’ fire scars per month, along with an
approximately equal number of unburned areas suffering
anomalous effects that caused rapid NIR decreases, was
selected by visual analysis of pre- and post-fire VGT S-10
products. Examples of a pre- and post-fire training data
subsets are shown in Fig. 3a and b, and analysis of these
training data suggested that discrimination of ‘true’ from
‘false’ fire scars (mapped with the NIR decrease criteria) could be improved by using the VGT red and SWIR
channels, along with two ratio-based indices; the NDVI
and the shortwave infrared vegetation index (SWVI):

where qNIR, qred, and qSWIR are the spectral reflectance
recorded in the VGT NIR, red, andSWIR bands, respectively.
A series of additional threshold-based rules using preand
post-fire red, SWIR, NDVI, and SWVI data were
therefore generated to minimise commission errors with
regard to burned area detection. In defining these rules, we
noted that NDVI has previously been used to investigate
monthly and annual burned area in various forest and
grassland environments (Fredriksen, Lanaas, & Mbaye,
1990; Kasischke & French, 1995; Kasischke et al., 1993;
Li, Nadon, & Cihlar, 2000; Martin & Chuvieco, 1993),
whilst Fraser & Li (2002) found SWVI chnages between
the start- and end-of-season to be an even stronger
discriminator of burned forest in the Canadian boreal
region. We did not utilise the VGT blue band because
this provides little useful information on burned area
(Fraser, Li, & Landry, 2000), but rather provides data of
use to various atmospheric correction procedures (Kaufman
& Tanre´, 1992).
Our rules for the detection of burned areas are required to
vary over the fire season because many of the phenomena
that can be incorrectly identified as burned areas are active
only at certain times, for example, melting of frost, ice, or
snow or the falling of leaves. Table 1 details the full set of
rules, and it should be noted that the threshold (Ti) used for
each rule is temporally varying, its value being determined
from the training data of that month. The training data were,
as far as possible, distributed to fully cover the geographic
locations and land cover classes encompassed in the Russian
Federation and for each ‘true’ fire scar contained within the
training data set the mean (M) and standard deviation (r) of
the reflectance and vegetation index values in both the preand
post-fire images were calculated to define the threshold
values. For Rules 1–5, the difference threshold Ti was
determined from (3):

For Rules 6–9, the absolute upper threshold Ti was
determined using (4), whilst for Rule 10, the absolute lower
threshold Ti was determined using (5):

where min and max denote the minimum and maximum
value over the entire training data set comprising true
burned areas.
0/5000
From: -
To: -
Results (Indonesian) 1: [Copy]
Copied!
In this way, the unburned areas aremore likely to remain as a constant cover type in themonthly composite data, rather than undergoing the vegetation–ice/frost–vegetation cycle seen in the individualS10 products. 4. Algorithm detailsAs discussed in Section 3, a significant fall in NIRreflectance between pre- and post-fire image dates isregarded as our primary method for mapping burned areaswith VGT. SWIR reflectance also decreases following a fireevent, but we note that this often rapidly recovers a fewweeks after burning, so that in some cases, only a smalldecrease in SWIR reflectance is observed between monthlypre- and post-fire SWIR data. This is demonstrated in Fig.2c and d, where a 36% fall in SWIR reflectance occursbetween the pre- and post-fire dates, compared to a 62% fallin the NIR reflectance.Although Fig. 2 shows that a clear and ubiquitous NIRreflectance decrease accompanies burning, because otherphenomena unrelated to fire also cause a similar NIRreflectance change we risk significant commission errorsif we rely on this criterion alone to identify burned areas.Such phenomena include the residual appearance of ice,snow, or frost in the pre-fire image composites, which hasmelted in the post-fire image, residual cloud shadows thatare present in the post-fire image but not the pre-fire image(i.e., partly shadowed pixels having erroneously passed theprior shadow-excluding processing) or the senescence ofvegetation and fall-off of leaves from deciduous trees inthe post-fire image. Further errors can occur due to thewide swath of the VGT sensor, which can lead to abnormallyhigh reflectance observations at large viewing zenithangles due to surface BRDF effects (Stroppiana et al.,2002). This may cause problems if the pre- and post-firedata of a pixel were obtained at very different viewinggeometries.Although the phenomena outlined above may cause NIRdecreases similar to that resulting from fire, further spectralanalysis can help in discriminating ‘true’ burnt areas fromthese other effects. A training data set consisting of aminimum of 15 ‘true’ fire scars per month, along with anapproximately equal number of unburned areas sufferinganomalous effects that caused rapid NIR decreases, wasselected by visual analysis of pre- and post-fire VGT S-10products. Examples of a pre- and post-fire training datasubsets are shown in Fig. 3a and b, and analysis of thesetraining data suggested that discrimination of ‘true’ from‘false’ fire scars (mapped with the NIR decrease criteria) could be improved by using the VGT red and SWIRchannels, along with two ratio-based indices; the NDVIand the shortwave infrared vegetation index (SWVI):where qNIR, qred, and qSWIR are the spectral reflectancerecorded in the VGT NIR, red, andSWIR bands, respectively.A series of additional threshold-based rules using preandpost-fire red, SWIR, NDVI, and SWVI data weretherefore generated to minimise commission errors withregard to burned area detection. In defining these rules, wenoted that NDVI has previously been used to investigatemonthly and annual burned area in various forest andgrassland environments (Fredriksen, Lanaas, & Mbaye,1990; Kasischke & French, 1995; Kasischke et al., 1993;Li, Nadon, & Cihlar, 2000; Martin & Chuvieco, 1993),whilst Fraser & Li (2002) found SWVI chnages betweenthe start- and end-of-season to be an even strongerdiscriminator of burned forest in the Canadian borealregion. We did not utilise the VGT blue band becausethis provides little useful information on burned area(Fraser, Li, & Landry, 2000), but rather provides data ofuse to various atmospheric correction procedures (Kaufman& Tanre´, 1992).Our rules for the detection of burned areas are required tovary over the fire season because many of the phenomenathat can be incorrectly identified as burned areas are activeonly at certain times, for example, melting of frost, ice, orsnow or the falling of leaves. Table 1 details the full set ofrules, and it should be noted that the threshold (Ti) used foreach rule is temporally varying, its value being determinedfrom the training data of that month. The training data were,as far as possible, distributed to fully cover the geographiclocations and land cover classes encompassed in the RussianFederation and for each ‘true’ fire scar contained within thetraining data set the mean (M) and standard deviation (r) ofthe reflectance and vegetation index values in both the preandpost-fire images were calculated to define the thresholdvalues. For Rules 1–5, the difference threshold Ti wasdetermined from (3):For Rules 6–9, the absolute upper threshold Ti wasdetermined using (4), whilst for Rule 10, the absolute lowerthreshold Ti was determined using (5):where min and max denote the minimum and maximumvalue over the entire training data set comprising trueburned areas.
Being translated, please wait..
Results (Indonesian) 2:[Copy]
Copied!
Dengan cara ini, daerah yang terbakar yang
lebih mungkin untuk tetap sebagai jenis penutup konstan dalam
data yang komposit bulanan, daripada menjalani vegetation-
siklus es / es-vegetasi terlihat di masing-masing
produk S10.
4. Rincian algoritma
Sebagaimana dibahas dalam Bagian 3, penurunan yang signifikan dalam NIR
reflektansi antara tanggal gambar pra dan pasca-kebakaran
dianggap sebagai metode utama kami untuk pemetaan dibakar daerah
dengan VGT. SWIR reflektansi juga menurun setelah kebakaran
acara, tapi kami mencatat bahwa ini sering cepat pulih beberapa
minggu setelah pembakaran, sehingga dalam beberapa kasus, hanya kecil
penurunan SWIR reflektansi diamati antara bulanan
Data SWIR pra dan pasca kebakaran. Hal ini ditunjukkan pada Gambar.
2c dan d, di mana 36% penurunan SWIR pantulan terjadi
antara sebelum dan sesudah kebakaran tanggal, dibandingkan dengan penurunan 62%
dalam reflektansi NIR.
Meskipun Gambar. 2 menunjukkan bahwa NIR jelas dan di mana-mana
penurunan reflektansi menyertai terbakar, karena lain
fenomena yang tidak terkait dengan api juga menyebabkan NIR setara
perubahan reflektansi kita berisiko kesalahan komisi yang signifikan
jika kita mengandalkan kriteria ini saja untuk mengidentifikasi daerah-daerah yang terbakar.
Fenomena tersebut termasuk penampilan sisa es,
salju, atau es di komposit gambar pre-api, yang telah
meleleh di gambar pasca-kebakaran, bayangan awan residual yang
hadir pada gambar pasca-api tapi tidak gambar pre-api
(yaitu, piksel sebagian dibayangi memiliki keliru lulus
sebelum bayangan-tidak termasuk pengolahan) atau penuaan dari
vegetasi dan jatuh-off daun dari pohon gugur di
gambar pasca-api. Kesalahan lebih lanjut dapat terjadi karena
petak luas dari sensor VGT, yang dapat menyebabkan abnormal
pengamatan reflektansi tinggi pada melihat zenith besar
sudut karena permukaan efek BRDF (Stroppiana et al.,
2002). Hal ini dapat menyebabkan masalah jika sebelum dan sesudah kebakaran
data piksel diperoleh pada pandang yang sangat berbeda
geometri.
Meskipun fenomena yang diuraikan di atas dapat menyebabkan NIR
menurun mirip dengan yang dihasilkan dari api, spektral lanjut
analisis dapat membantu dalam membedakan 'benar' kawasan yang terbakar dari
efek lainnya. Satu set data training yang terdiri dari
minimal 15 'benar' bekas kebakaran per bulan, bersama dengan
jumlah kurang lebih sama dari daerah yang tidak terbakar menderita
efek anomali yang menyebabkan cepat NIR menurun, itu
dipilih oleh analisis visual sebelum dan sesudah api VGT S -10
produk. Contoh dari pra dan pasca-kebakaran data training
subset ditunjukkan pada Gambar. 3a dan b, dan analisis ini
data training menyarankan bahwa diskriminasi dari 'benar' dari
'palsu' bekas api (dipetakan dengan kriteria penurunan NIR) dapat ditingkatkan dengan menggunakan VGT merah dan SWIR
saluran, bersama dengan dua indeks berdasarkan rasio- ; NDVI
dan gelombang pendek indeks vegetasi inframerah (SWVI): mana qNIR, qred, dan qSWIR adalah reflektansi spektral dicatat dalam VGT NIR, merah, band andSWIR, masing-masing. Serangkaian aturan berbasis threshold tambahan menggunakan preand merah pasca-api , Data SWIR, NDVI, dan SWVI yang karena dihasilkan untuk meminimalkan kesalahan komisi dengan hal deteksi daerah dibakar. Dalam mendefinisikan aturan-aturan ini, kami mencatat bahwa NDVI sebelumnya telah digunakan untuk menyelidiki daerah yang terbakar bulanan dan tahunan di berbagai hutan dan lingkungan padang rumput (Fredriksen, Lanaas, & Mbaye, 1990; Kasischke & French, 1995; Kasischke et al, 1993;. Li , Nadon, & Cihlar, 2000; Martin & Chuvieco, 1993), sementara Fraser & Li (2002) menemukan SWVI chnages antara para start dan end-of-musim menjadi lebih kuat discriminator hutan terbakar di boreal Kanada wilayah. Kami tidak menggunakan VGT pita biru karena ini memberikan sedikit informasi yang berguna di daerah terbakar (Fraser, Li, & Landry, 2000), melainkan memberikan data penggunaan untuk berbagai prosedur koreksi atmosfer (Kaufman & Tanre', 1992). Aturan kami untuk mendeteksi daerah terbakar yang diperlukan untuk bervariasi selama musim kebakaran karena banyak fenomena yang bisa salah diidentifikasi sebagai daerah yang terbakar aktif hanya pada waktu tertentu, misalnya, pencairan es, es, atau salju atau jatuh dari daun . Tabel 1 Rincian set lengkap aturan, dan perlu dicatat bahwa ambang batas (Ti) digunakan untuk setiap aturan yang beragam secara temporal, nilainya ditentukan dari data pelatihan bulan itu. Data pelatihan yang, sejauh mungkin, didistribusikan untuk sepenuhnya menutupi geografis lokasi dan kelas tutupan lahan dicakup dalam Rusia Federasi dan untuk setiap 'benar' bekas kebakaran yang terkandung dalam data training set mean (M) dan standar deviasi (r ) dari reflektansi dan vegetasi nilai indeks di kedua preand gambar pasca-kebakaran dihitung untuk menentukan ambang batas nilai. Untuk Aturan 1-5, ambang perbedaan Ti itu ditentukan dari (3): Untuk Aturan 6-9, batas atas mutlak Ti itu ditentukan dengan (4), sementara untuk Rule 10, mutlak lebih rendah ambang Ti ditentukan dengan menggunakan (5 ): di mana min dan max menyatakan minimum dan maksimum nilai atas seluruh data set training yang terdiri benar daerah yang terbakar.











































Being translated, please wait..
 
Other languages
The translation tool support: Afrikaans, Albanian, Amharic, Arabic, Armenian, Azerbaijani, Basque, Belarusian, Bengali, Bosnian, Bulgarian, Catalan, Cebuano, Chichewa, Chinese, Chinese Traditional, Corsican, Croatian, Czech, Danish, Detect language, Dutch, English, Esperanto, Estonian, Filipino, Finnish, French, Frisian, Galician, Georgian, German, Greek, Gujarati, Haitian Creole, Hausa, Hawaiian, Hebrew, Hindi, Hmong, Hungarian, Icelandic, Igbo, Indonesian, Irish, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Kinyarwanda, Klingon, Korean, Kurdish (Kurmanji), Kyrgyz, Lao, Latin, Latvian, Lithuanian, Luxembourgish, Macedonian, Malagasy, Malay, Malayalam, Maltese, Maori, Marathi, Mongolian, Myanmar (Burmese), Nepali, Norwegian, Odia (Oriya), Pashto, Persian, Polish, Portuguese, Punjabi, Romanian, Russian, Samoan, Scots Gaelic, Serbian, Sesotho, Shona, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Sundanese, Swahili, Swedish, Tajik, Tamil, Tatar, Telugu, Thai, Turkish, Turkmen, Ukrainian, Urdu, Uyghur, Uzbek, Vietnamese, Welsh, Xhosa, Yiddish, Yoruba, Zulu, Language translation.

Copyright ©2024 I Love Translation. All reserved.

E-mail: