-
Notifications
You must be signed in to change notification settings - Fork 0
/
pyeo_sepal.ini
170 lines (134 loc) · 7.34 KB
/
pyeo_sepal.ini
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
# Edit the variables below and save before running the pipeline
[run_mode]
# flag for linear or parallel processing of the raster and vector processing pipeline
do_parallel = False
# wall_time_hours and qsub_processor_options are currently hardcoded to 24 hours, at line 495 of acd_national.py
qsub_processor_options = nodes=1:ppn=16,vmem=64Gb
wall_time_hours = 3
watch_time_hours = 3
watch_period_seconds = 60
# needs to be in the format "nodes=1:ppn=16,vmem=64Gb"
[forest_sentinel]
# aoi_name: The name of this area of interest. No spaces.
# aoi_name=Kenya
# Acquisition dates in the form yyyymmdd
start_date=20230101
end_date=20230210
# Dates to download and preprocess for the initial cloud-free composite
composite_start=20220101
composite_end=20221231
# EPSG code for Kenya - north of equator and east of 36°E is EPSG:21097
# See https://epsg.io/21097 and https://spatialreference.org/ref/epsg/21097/
epsg=21097
# Cloud cover threshold for imagery to download
cloud_cover=25
# Certainty value above which a pixel is considered a cloud from sen2cor
cloud_certainty_threshold=0
# path to the trained machine learning model for land cover in Kenya
model= ./models/model_36MYE_Unoptimised_20230505_no_haze.pkl
[environment]
# pyeo_dir needs to be an absolute path
pyeo_dir = /home/sepal-user/20230626_pyeo_installation/pyeo_1
# tile_dir needs to be an absolute path
tile_dir = /home/sepal-user/20230626_pyeo_installation
# Relative paths are relative to pyeo_dir\
integrated_dir = ./integrated
roi_dir = ./roi
roi_filename = kfs_roi_subset_c.shp
geometry_dir = ./geometry
s2_tiles_filename = kenya_s2_tiles.shp
log_dir = ./log
log_filename = sepal_venv.log
credentials_path = ./credentials/credentials.ini
environment_manager = venv
# conda_environment
#conda_directory = /home/m/mp730/miniconda3
#conda_env_name = pyeo_env
# Path to the sen2cor preprocessor script, L2A_Process. Usually in the bin/ folder of your sen2cor installation.
sen2cor_path = /usr/local/lib/Sen2Cor-02.10.01-Linux64/bin/L2A_Process
[raster_processing_parameters]
# ***** STEP 1 ***** MOVE STEP 1 TO ITS OWN SECTION [] BEFORE [RASTER PROCESSING PARAMETERS] AS AN INDEPENDENT STAGE
do_tile_intersection = True
# **************************************************************************************************************************
# ***** STEP 2 SETUP GENERAL RASTER PROCESSING OPTIONS *****
# ***** NOTE: MOVE [forest_sentinel] OPTIONS TO APPROPRIATE SECTION BELOW ... DATE RANGES, CLOUD, ETC.*****
do_raster = True
chunks = 10
do_skip_existing = True
do_quicklooks = True
do_delete = False
do_zip = False
# do_zip, consider removing
do_all = False
# do_dev, consider removing
do_update = False
# do_update, consider removing
do_dev = True
# **************************************************************************************************************************
# ***** STEP 3 SETUP GENERAL RASTER PROCESSING PARAMETERS *****
# download_source = scihub
download_source = dataspace
# granules below this size in MB will not be downloaded
faulty_granule_threshold = 350
# list of strings with the band name elements of the image file names in "" string notation
# the wavebands specified here must match those used to build the random forest model specified in the Classify section below
band_names = ["B02", "B03", "B04", "B08"]
# file name pattern to search for when identifying band file locations in "" string notation
resolution_string = "10m"
# spatial resolution of the output raster files in metres. Can be any resolution, not just 10, 20 or 60 as in the default band resolutions of Sentinel-2
output_resolution = 10
# **************************************************************************************************************************
# ***** STEP 4 DOWNLOAD REFERENCE IMAGES AND BUILD A MEDIAN COMPOSITE *****
do_build_composite = True
# set buffer in number of pixels for dilating the SCL cloud mask (recommend 10 pixels of 10 m) for the composite building
buffer_size_cloud_masking_composite = 10
# maximum number of images to be downloaded for compositing, in order of least cloud cover
download_limit = 5
# **************************************************************************************************************************
# ***** STEP 5: DOWNLOAD CHANGE DETECTION IMAGES FOR THE REQUIRED DATE RANGE *****
do_download = True
# set buffer in number of pixels for dilating the SCL cloud mask (recommend 30 pixels of 10 m) for the change detection
buffer_size_cloud_masking = 20
# **************************************************************************************************************************
# ***** STEP 6: CLASSIFIY THE COMPOSITE AND CHANGE DETECTION IMAGES *****
# ***** NOTE: MOVE MODEL SPECIFICATION TO THIS SECTION*****
do_build_prob_image = False
# do_build_prob_image, consider removing
do_classify = True
# list of strings with class labels starting from class 1. Must match the trained model that was used.
class_labels = ["primary forest", "plantation forest", "bare soil", "crops", "grassland", "open water", "burn scar", "cloud", "cloud shadow", "haze", "sparse woodland", "dense woodland", "artificial"]
# if sieve is 0, no sieve is applied. If >0, the classification images will be sieved using gdal and all contiguous groups of pixels smaller than this number will be eliminated
sieve = 0
# **************************************************************************************************************************
# ***** STEP 7: DETECT CHANGES AND BUILD RASTER REPORTS *****
# ***** NOTE: ADD REPORT BINARISATION PARAMETERS TO THIS SECTION MIN_DETECTION; CONSISTENCY ETC.*****
do_change = True
# find subsequent changes from any of these classes. Must match the trained model that was used.
change_from_classes = [1, 2]
# to any of these classes. Must match the trained model that was used.
change_to_classes = [3]
# **************************************************************************************************************************
# ***** STEP 8: VECTOR ANALYSIS OF TILE RASTER REPORTS *****
[vector_processing_parameters]
level_1_filename = gadm41_KEN_1.json
# vectorisation currently hardcoded to use level_1_filename
level_2_filename = gadm41_KEN_2.json
level_3_filename = gadm41_KEN_3.json
do_delete_existing_vector = True
do_vectorise = True
# ***** STEP 9: INTEGRATE VECTOR ANALYSES TO NATIONAL SCOPE *****
do_integrate = True
# **************************************************************************************************************************
# ***** STEP 10: FILTER NATIONAL SCOPE VECTORISED FOREST ALERTS *****
do_filter = False
# If there are any strings within counties_of_interest list, filtering by county will be attempted
counties_of_interest = []
# Counties_of_interest = ["Kwale", "TransNzoia"]
minimum_area_to_report_m2 = 120
# **************************************************************************************************************************
# ***** STEP 11: FILTER NATIONAL SCOPE VECTORISED FOREST ALERTS *****
# MANUAL ASSESSMENT OF ALERTS BY KFS STAFF AND FLAGGING FOR ALERTS FOR IN-PERSON INVESTIGATION BY RANGERS
# **************************************************************************************************************************
# ***** STEP 12: AUTOMATED DISTRIBUTION OF FILTERED ALERTS BY MESSAGING (E.G. WHATSAPP) *****
do_distribution = False
# **************************************************************************************************************************