Qualche anno fa analizzai i dati del progetto CORINE Land Cover relativi alla Campania, focalizzandomi sui tre censimenti tra il 2000 ed il 2012. Oggi invece estenderò quell'analisi inserendo sia i dati del 1990 che gli ultimi disponibili attualmente, cioè quelli del 2018.

L'analisi questa volta la farò usando Python e la estenderò al resto dell'Italia verso la fine.

Prima di iniziare

Librerie

Per raggiungere l'obiettivo userò le librerie che seguono:

from pathlib import Path
import geopandas as gpd
import pandas as pd
from functools import reduce
import matplotlib.pyplot as plt
from matplotlib.gridspec import GridSpec

import warnings
warnings.filterwarnings('ignore')
warnings.simplefilter('ignore')

Fonti dati

Ho rielaborato i dati CORINE Land Cover dal 1990 al 2018 inserendoli in un unico GeoPackage che è possibile scaricare da qui. I dati sono stati processati preventivamente usando PostGIS in modo da associare ai poligoni della CLC anche quelli dei Comuni italiani al 2022 dell'ISTAT. E' stato indispensabile questo passaggio preventivo perchè ho dovuto correggere gli errori topologici sui dati ISTAT e, al momento, trovo molto più comodo correggere gli errori topologici tramite SQL piuttosto che con Python.

main_folder = Path('/home/max/Downloads/clcitalia')
dataset = main_folder.joinpath('clcitalia.gpkg')

Contenuti

• 1. Lettura e normalizzazione dei dati
• 2. Focus sulla Campania
• 3. Analisi a scala nazionale
• Conclusione

1. Lettura e normalizzazione dei dati

In questa fase ho proceduto ad estrarre un po' di dati statistici che userò nei passaggi successivi. Sono perlopiu analisi mirate al calcolo delle superfici che ho espresso in ettari. Fatte queste analisi mi sono liberato dei dati geometrici perchè, non dovendoli più usare, mi alleggeriscono le analisi successive in termini di prestazioni hardware.

%%time

# CLC 1990
clc1990 = gpd.read_file(dataset, layer='clc1990')
clc1990.insert(loc=10, column='sup_com_ha', value=clc1990.sup_com_mq / 10000)
clc1990.insert(loc=11, column='surface_ha_1990', value=round(clc1990.geometry.area / 10000, 4))
clc1990.drop(columns={'sup_com_mq', 'geometry'}, inplace=True)

clc1990.head(10)

CPU times: user 1min 4s, sys: 687 ms, total: 1min 4s
Wall time: 1min 4s

	code_90	label1	label2	label3	pro_com	comune	provincia	regione	sup_com_ha	surface_ha_1990
0	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.1386
1	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	276.7171
2	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	35.4001
3	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	2.2285
4	241	Agricultural areas	Heterogeneous agricultural areas	Annual crops associated with permanent crops	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.3860
5	242	Agricultural areas	Heterogeneous agricultural areas	Complex cultivation patterns	101005.0	Castelsilano	Crotone	Calabria	4006.296249	8.6550
6	243	Agricultural areas	Heterogeneous agricultural areas	Land principally occupied by agriculture, with...	101005.0	Castelsilano	Crotone	Calabria	4006.296249	37.2054
7	243	Agricultural areas	Heterogeneous agricultural areas	Land principally occupied by agriculture, with...	101005.0	Castelsilano	Crotone	Calabria	4006.296249	114.5143
8	311	Forest and semi natural areas	Forests	Broad-leaved forest	101005.0	Castelsilano	Crotone	Calabria	4006.296249	44.3502
9	311	Forest and semi natural areas	Forests	Broad-leaved forest	101005.0	Castelsilano	Crotone	Calabria	4006.296249	32.6148

%%time

# CLC 2000
clc2000 = gpd.read_file(dataset, layer='clc2000')
clc2000.insert(loc=10, column='sup_com_ha', value=clc2000.sup_com_mq / 10000)
clc2000.insert(loc=11, column='surface_ha_2000', value=round(clc2000.geometry.area / 10000, 4))
clc2000.drop(columns={'sup_com_mq', 'geometry'}, inplace=True)

clc2000.head(10)

CPU times: user 1min 14s, sys: 388 ms, total: 1min 14s
Wall time: 1min 14s

	code_00	label1	label2	label3	pro_com	comune	provincia	regione	sup_com_ha	surface_ha_2000
0	321	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Natural grasslands	2170.0	Alto Sermenza	Vercelli	Piemonte	6032.586221	11.9494
1	321	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Natural grasslands	2170.0	Alto Sermenza	Vercelli	Piemonte	6032.586221	42.8232
2	321	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Natural grasslands	2170.0	Alto Sermenza	Vercelli	Piemonte	6032.586221	279.4458
3	321	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Natural grasslands	2170.0	Alto Sermenza	Vercelli	Piemonte	6032.586221	120.3762
4	321	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Natural grasslands	2170.0	Alto Sermenza	Vercelli	Piemonte	6032.586221	80.1545
5	311	Forest and semi natural areas	Forests	Broad-leaved forest	46008.0	Careggine	Lucca	Toscana	2408.101499	1664.3039
6	324	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Transitional woodland-shrub	46008.0	Careggine	Lucca	Toscana	2408.101499	25.8278
7	324	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Transitional woodland-shrub	46008.0	Careggine	Lucca	Toscana	2408.101499	73.5182
8	333	Forest and semi natural areas	Open spaces with little or no vegetation	Sparsely vegetated areas	46008.0	Careggine	Lucca	Toscana	2408.101499	34.7050
9	333	Forest and semi natural areas	Open spaces with little or no vegetation	Sparsely vegetated areas	46008.0	Careggine	Lucca	Toscana	2408.101499	101.5072

%%time

# CLC 2006
clc2006 = gpd.read_file(dataset, layer='clc2006')
clc2006.insert(loc=10, column='sup_com_ha', value=clc2006.sup_com_mq / 10000)
clc2006.insert(loc=11, column='surface_ha_2006', value=round(clc2006.geometry.area / 10000, 4))
clc2006.drop(columns={'sup_com_mq', 'geometry'}, inplace=True)

clc2006.head(10)

CPU times: user 1min 10s, sys: 328 ms, total: 1min 11s
Wall time: 1min 11s

	code_06	label1	label2	label3	pro_com	comune	provincia	regione	sup_com_ha	surface_ha_2006
0	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	99.9449
1	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	161.6097
2	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	35.3999
3	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.1332
4	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	5.0278
5	241	Agricultural areas	Heterogeneous agricultural areas	Annual crops associated with permanent crops	101005.0	Castelsilano	Crotone	Calabria	4006.296249	2.8168
6	242	Agricultural areas	Heterogeneous agricultural areas	Complex cultivation patterns	101005.0	Castelsilano	Crotone	Calabria	4006.296249	0.1570
7	242	Agricultural areas	Heterogeneous agricultural areas	Complex cultivation patterns	101005.0	Castelsilano	Crotone	Calabria	4006.296249	165.7342
8	243	Agricultural areas	Heterogeneous agricultural areas	Land principally occupied by agriculture, with...	101005.0	Castelsilano	Crotone	Calabria	4006.296249	36.2394
9	311	Forest and semi natural areas	Forests	Broad-leaved forest	101005.0	Castelsilano	Crotone	Calabria	4006.296249	32.6261

%%time

# CLC 2012
clc2012 = gpd.read_file(dataset, layer='clc2012')
clc2012.insert(loc=10, column='sup_com_ha', value=clc2012.sup_com_mq / 10000)
clc2012.insert(loc=11, column='surface_ha_2012', value=round(clc2012.geometry.area / 10000, 4))
clc2012.drop(columns={'sup_com_mq', 'geometry'}, inplace=True)

clc2012.head(10)

CPU times: user 1min 20s, sys: 588 ms, total: 1min 21s
Wall time: 1min 21s

	code_12	label1	label2	label3	pro_com	comune	provincia	regione	sup_com_ha	surface_ha_2012
0	243	Agricultural areas	Heterogeneous agricultural areas	Land principally occupied by agriculture, with...	101005.0	Castelsilano	Crotone	Calabria	4006.296249	0.5968
1	313	Forest and semi natural areas	Forests	Mixed forest	101005.0	Castelsilano	Crotone	Calabria	4006.296249	187.4011
2	324	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Transitional woodland-shrub	101005.0	Castelsilano	Crotone	Calabria	4006.296249	29.7040
3	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	99.9449
4	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	161.6097
5	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	420.1994
6	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	35.3999
7	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.1332
8	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	5.0278
9	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.6430

%%time

# CLC 2018
clc2018 = gpd.read_file(dataset, layer='clc2018')
clc2018.insert(loc=10, column='sup_com_ha', value=clc2018.sup_com_mq / 10000)
clc2018.insert(loc=11, column='surface_ha_2018', value=round(clc2018.geometry.area / 10000, 4))
clc2018.drop(columns={'sup_com_mq', 'geometry'}, inplace=True)

clc2018.head(10)

CPU times: user 1min 12s, sys: 560 ms, total: 1min 12s
Wall time: 1min 12s

	code_18	label1	label2	label3	pro_com	comune	provincia	regione	sup_com_ha	surface_ha_2018
0	242	Agricultural areas	Heterogeneous agricultural areas	Complex cultivation patterns	101005.0	Castelsilano	Crotone	Calabria	4006.296249	0.1570
1	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	161.6097
2	211	Agricultural areas	Arable land	Non-irrigated arable land	101005.0	Castelsilano	Crotone	Calabria	4006.296249	420.1994
3	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	5.0278
4	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.6430
5	223	Agricultural areas	Permanent crops	Olive groves	101005.0	Castelsilano	Crotone	Calabria	4006.296249	34.7345
6	241	Agricultural areas	Heterogeneous agricultural areas	Annual crops associated with permanent crops	101005.0	Castelsilano	Crotone	Calabria	4006.296249	2.8168
7	241	Agricultural areas	Heterogeneous agricultural areas	Annual crops associated with permanent crops	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.3899
8	241	Agricultural areas	Heterogeneous agricultural areas	Annual crops associated with permanent crops	101005.0	Castelsilano	Crotone	Calabria	4006.296249	13.0463
9	243	Agricultural areas	Heterogeneous agricultural areas	Land principally occupied by agriculture, with...	101005.0	Castelsilano	Crotone	Calabria	4006.296249	1.1525

Le colonne label* fanno riferimento alla nomenclatura delle tre classi della mappa CORINE Land Cover per ogni singolo anno. Per interpretare la nomenclatura vai qui. Come puoi vedere sono presenti le colonne relative al Comune, Provincia e Regione di apparteneza, l'estensione dell'area comunale(sup_com_ha) e quelle relative alle superfici della CLC che mi sono calcolato in precedenza.

2. Focus sulla Campania

Per appartenza territoriale partirò proprio dalla Campania, lo faccio anche per riprendere parte della analisi del mio vecchio articolo.

clc1990_campania = clc1990[clc1990['regione'] == 'Campania']
clc2000_campania = clc2000[clc2000['regione'] == 'Campania']
clc2006_campania = clc2006[clc2006['regione'] == 'Campania']
clc2012_campania = clc2012[clc2012['regione'] == 'Campania']
clc2018_campania = clc2018[clc2018['regione'] == 'Campania']

Test di visualizzazione

clc2000_campania

	code_00	label1	label2	label3	pro_com	comune	provincia	regione	sup_com_ha	surface_ha_2000
14	111	Artificial surfaces	Urban fabric	Continuous urban fabric	65134.0	Sapri	Salerno	Campania	1420.007490	82.5207
15	112	Artificial surfaces	Urban fabric	Discontinuous urban fabric	65134.0	Sapri	Salerno	Campania	1420.007490	36.5516
16	241	Agricultural areas	Heterogeneous agricultural areas	Annual crops associated with permanent crops	65134.0	Sapri	Salerno	Campania	1420.007490	17.7890
17	242	Agricultural areas	Heterogeneous agricultural areas	Complex cultivation patterns	65134.0	Sapri	Salerno	Campania	1420.007490	0.6866
18	242	Agricultural areas	Heterogeneous agricultural areas	Complex cultivation patterns	65134.0	Sapri	Salerno	Campania	1420.007490	56.9832
...	...	...	...	...	...	...	...	...	...	...
177542	323	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Sclerophyllous vegetation	61072.0	Rocchetta e Croce	Caserta	Campania	1300.547184	206.0290
177543	324	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Transitional woodland-shrub	61072.0	Rocchetta e Croce	Caserta	Campania	1300.547184	2.3093
177544	221	Agricultural areas	Permanent crops	Vineyards	61072.0	Rocchetta e Croce	Caserta	Campania	1300.547184	4.0201
177545	321	Forest and semi natural areas	Scrub and/or herbaceous vegetation associations	Natural grasslands	61072.0	Rocchetta e Croce	Caserta	Campania	1300.547184	1.8517
177546	222	Agricultural areas	Permanent crops	Fruit trees and berry plantations	61072.0	Rocchetta e Croce	Caserta	Campania	1300.547184	1.9998

11361 rows × 10 columns

2.1.1 Stato al 1990

Come primo step mi sono concentrato sui dati del 1990; ho pulito un po' la tabella per rendere meglio comprensibile la sua lettura ed ho estratto valori per tutte e tre le classi. Quello che ho fatto per il 1990 l'ho ripetuto per i restanti anni.

Livello 1

campania_clc1990_label1 = pd.DataFrame(clc1990_campania.groupby('label1')['surface_ha_1990'].sum())
campania_clc1990_label1.reset_index(inplace=True)
campania_clc1990_label1.rename(columns={'label1': 'CORINE Land Cover - level 1'}, inplace=True)
campania_clc1990_label1

	CORINE Land Cover - level 1	surface_ha_1990
0	Agricultural areas	761041.0303
1	Artificial surfaces	80649.9550
2	Forest and semi natural areas	522870.7142
3	Water bodies	1593.6569
4	Wetlands	608.8997

Livello 2

campania_clc1990_label2 = pd.DataFrame(clc1990_campania.groupby('label2')['surface_ha_1990'].sum())
campania_clc1990_label2.reset_index(inplace=True)
campania_clc1990_label2.rename(columns={'label2': 'CORINE Land Cover - level 2'}, inplace=True)
campania_clc1990_label2

	CORINE Land Cover - level 2	surface_ha_1990
0	Arable land	330066.4318
1	Artificial, non-agricultural vegetated areas	1046.0670
2	Forests	383826.2982
3	Heterogeneous agricultural areas	297503.9919
4	Industrial, commercial and transport units	8083.1341
5	Inland waters	1567.8002
6	Inland wetlands	608.8997
7	Marine waters	25.8567
8	Mine, dump and construction sites	1631.6247
9	Open spaces with little or no vegetation	8966.3806
10	Pastures	15620.1214
11	Permanent crops	117850.4852
12	Scrub and/or herbaceous vegetation associations	130078.0354
13	Urban fabric	69889.1292

Livello 3

campania_clc1990_label3 = pd.DataFrame(clc1990_campania.groupby('label3')['surface_ha_1990'].sum())
campania_clc1990_label3.reset_index(inplace=True)
campania_clc1990_label3.rename(columns={'label3': 'CORINE Land Cover - level 3'}, inplace=True)
campania_clc1990_label3

	CORINE Land Cover - level 3	surface_ha_1990
0	Agro-forestry areas	420.5049
1	Airports	383.8998
2	Annual crops associated with permanent crops	42881.8119
3	Bare rocks	1122.7257
4	Beaches, dunes, sands	2282.0112
5	Broad-leaved forest	367521.2801
6	Coastal lagoons	25.8567
7	Complex cultivation patterns	165047.1097
8	Coniferous forest	7457.8770
9	Construction sites	25.4060
10	Continuous urban fabric	32608.0637
11	Discontinuous urban fabric	37281.0655
12	Dump sites	207.8811
13	Fruit trees and berry plantations	56044.2433
14	Green urban areas	731.9919
15	Industrial or commercial units	6590.0765
16	Inland marshes	608.8997
17	Land principally occupied by agriculture, with...	89154.5654
18	Mineral extraction sites	1398.3376
19	Mixed forest	8847.1411
20	Moors and heathland	16.7825
21	Natural grasslands	57150.7848
22	Non-irrigated arable land	304750.4005
23	Olive groves	58951.7577
24	Pastures	15620.1214
25	Permanently irrigated land	25316.0313
26	Port areas	458.6026
27	Road and rail networks and associated land	650.5552
28	Sclerophyllous vegetation	19255.7695
29	Sparsely vegetated areas	5561.6437
30	Sport and leisure facilities	314.0751
31	Transitional woodland-shrub	53654.6986
32	Vineyards	2854.4842
33	Water bodies	1095.5920
34	Water courses	472.2082

2.1.2 Stato al 2000

campania_clc2000_label1 = pd.DataFrame(clc2000_campania.groupby('label1')['surface_ha_2000'].sum())
campania_clc2000_label1.reset_index(inplace=True)
campania_clc2000_label1.rename(columns={'label1': 'CORINE Land Cover - level 1'}, inplace=True)

campania_clc2000_label2 = pd.DataFrame(clc2000_campania.groupby('label2')['surface_ha_2000'].sum())
campania_clc2000_label2.reset_index(inplace=True)
campania_clc2000_label2.rename(columns={'label2': 'CORINE Land Cover - level 2'}, inplace=True)

campania_clc2000_label3 = pd.DataFrame(clc2000_campania.groupby('label3')['surface_ha_2000'].sum())
campania_clc2000_label3.reset_index(inplace=True)
campania_clc2000_label3.rename(columns={'label3': 'CORINE Land Cover - level 3'}, inplace=True)

2.1.3 Stato al 2006

campania_clc2006_label1 = pd.DataFrame(clc2006_campania.groupby('label1')['surface_ha_2006'].sum())
campania_clc2006_label1.reset_index(inplace=True)
campania_clc2006_label1.rename(columns={'label1': 'CORINE Land Cover - level 1'}, inplace=True)

campania_clc2006_label2 = pd.DataFrame(clc2006_campania.groupby('label2')['surface_ha_2006'].sum())
campania_clc2006_label2.reset_index(inplace=True)
campania_clc2006_label2.rename(columns={'label2': 'CORINE Land Cover - level 2'}, inplace=True)

campania_clc2006_label3 = pd.DataFrame(clc2006_campania.groupby('label3')['surface_ha_2006'].sum())
campania_clc2006_label3.reset_index(inplace=True)
campania_clc2006_label3.rename(columns={'label3': 'CORINE Land Cover - level 3'}, inplace=True)

2.1.4 Stato al 2012

campania_clc2012_label1 = pd.DataFrame(clc2012_campania.groupby('label1')['surface_ha_2012'].sum())
campania_clc2012_label1.reset_index(inplace=True)
campania_clc2012_label1.rename(columns={'label1': 'CORINE Land Cover - level 1'}, inplace=True)

campania_clc2012_label2 = pd.DataFrame(clc2012_campania.groupby('label2')['surface_ha_2012'].sum())
campania_clc2012_label2.reset_index(inplace=True)
campania_clc2012_label2.rename(columns={'label2': 'CORINE Land Cover - level 2'}, inplace=True)

campania_clc2012_label3 = pd.DataFrame(clc2012_campania.groupby('label3')['surface_ha_2012'].sum())
campania_clc2012_label3.reset_index(inplace=True)
campania_clc2012_label3.rename(columns={'label3': 'CORINE Land Cover - level 3'}, inplace=True)

2.1.5 Stato al 2018

campania_clc2018_label1 = pd.DataFrame(clc2018_campania.groupby('label1')['surface_ha_2018'].sum())
campania_clc2018_label1.reset_index(inplace=True)
campania_clc2018_label1.rename(columns={'label1': 'CORINE Land Cover - level 1'}, inplace=True)

campania_clc2018_label2 = pd.DataFrame(clc2018_campania.groupby('label2')['surface_ha_2018'].sum())
campania_clc2018_label2.reset_index(inplace=True)
campania_clc2018_label2.rename(columns={'label2': 'CORINE Land Cover - level 2'}, inplace=True)

campania_clc2018_label3 = pd.DataFrame(clc2018_campania.groupby('label3')['surface_ha_2018'].sum())
campania_clc2018_label3.reset_index(inplace=True)
campania_clc2018_label3.rename(columns={'label3': 'CORINE Land Cover - level 3'}, inplace=True)

2.2 Grafici

E' il momento di approfondire questi dati! Ho estratto grafici per tutte e tre le classi ed ho cercato di dare una interpretazione del perchè di alcuni fenomeni.

2.2.1 Grafici per il livello 1

lvl1 = [campania_clc1990_label1, campania_clc2000_label1, campania_clc2006_label1, campania_clc2012_label1, campania_clc2018_label1]
lvl1_df = reduce(lambda  left,right: pd.merge(left,right,on=['CORINE Land Cover - level 1'], how='outer'), lvl1)
lvl1_df.fillna(0, inplace=True)
lvl1_df.insert(loc=6, column='Surf.Diff. | 2018-1990', value=(lvl1_df.surface_ha_2018 - lvl1_df.surface_ha_1990))
lvl1_df.set_index('CORINE Land Cover - level 1', inplace=True)
lvl1_df.rename(columns={'surface_ha_1990': '1990', 'surface_ha_2000': '2000', 'surface_ha_2006': '2006', 'surface_ha_2012': '2012', 'surface_ha_2018': '2018'}, inplace=True)
lvl1_df.sort_index(ascending=True, inplace=True)
lvl1_df

	1990	2000	2006	2012	2018	Surf.Diff. | 2018-1990
CORINE Land Cover - level 1
Agricultural areas	761041.0303	753145.2349	755438.9748	752611.5454	752077.3861	-8963.6442
Artificial surfaces	80649.9550	89935.2388	97747.4059	101771.7643	102304.6984	21654.7434
Forest and semi natural areas	522870.7142	521005.1536	510731.5181	509096.9289	508050.0206	-14820.6936
Water bodies	1593.6569	2195.8983	2463.1760	2948.2571	2948.2571	1354.6002
Wetlands	608.8997	482.9084	386.6832	339.2620	339.2620	-269.6377

fig1 = plt.figure(figsize=(20, 10), dpi=100)
sub_fig = GridSpec(nrows=1, ncols=1, figure=fig1, hspace=0.35)
ax = fig1.add_subplot(sub_fig[0, 0])
lvl1_df.plot.bar(ax=ax, rot=0)

ax.yaxis.set_label_text('ha')
ax.set_xlabel('')
ax.grid(color='lightgrey', linestyle='dashdot')

plt.legend(loc='best')

<matplotlib.legend.Legend at 0x7f13b28b6da0>

E' visibile abbastanza chiaramente un netto aumento delle superfici artificiali con un conseguente arretramento di quelle forestate e di quelle agricole. Le aree forestate tra il 1990 ed il 2018 hanno subito un arretramento quasi doppio rispetto a quelle agricole e le superfici artificiali sono quelle che hanno maggiormente "beneficiato" di questo arretramento complessivo delle due classi di CLC. Si potrebbe dire che l'urbanizzato ha eroso foreste ed aree agricole in quasi 30 anni.

2.2.2 Grafici per il livello 2

lvl2 = [campania_clc1990_label2, campania_clc2000_label2, campania_clc2006_label2, campania_clc2012_label2, campania_clc2018_label2]
lvl2_df = reduce(lambda  left,right: pd.merge(left,right,on=['CORINE Land Cover - level 2'], how='outer'), lvl2)
lvl2_df.fillna(0, inplace=True)
lvl2_df.insert(loc=6, column='Surf.Diff. | 2000-1990', value=(lvl2_df.surface_ha_2000 - lvl2_df.surface_ha_1990))
lvl2_df.insert(loc=7, column='Surf.Diff. | 2006-2000', value=(lvl2_df.surface_ha_2006 - lvl2_df.surface_ha_2000))
lvl2_df.insert(loc=8, column='Surf.Diff. | 2012-2006', value=(lvl2_df.surface_ha_2012 - lvl2_df.surface_ha_2006))
lvl2_df.insert(loc=9, column='Surf.Diff. | 2018-2012', value=(lvl2_df.surface_ha_2018 - lvl2_df.surface_ha_2012))
lvl2_df.insert(loc=9, column='Surf.Diff. | 2018-1990', value=(lvl2_df.surface_ha_2018 - lvl2_df.surface_ha_1990))
lvl2_df.set_index('CORINE Land Cover - level 2', inplace=True)
lvl2_df = lvl2_df[['Surf.Diff. | 2000-1990', 'Surf.Diff. | 2006-2000', 'Surf.Diff. | 2012-2006', 'Surf.Diff. | 2018-2012', 'Surf.Diff. | 2018-1990']]
lvl2_df.sort_index(ascending=True, inplace=True)
lvl2_df

	Surf.Diff. | 2000-1990	Surf.Diff. | 2006-2000	Surf.Diff. | 2012-2006	Surf.Diff. | 2018-2012	Surf.Diff. | 2018-1990
CORINE Land Cover - level 2
Arable land	-6226.9986	-15007.4805	1362.7374	26.9417	-19844.8000
Artificial, non-agricultural vegetated areas	330.5367	-4.4113	58.5588	396.9458	781.6300
Forests	2963.0601	-4293.0050	1859.3801	-1391.9151	-862.4799
Heterogeneous agricultural areas	-947.6309	4546.7408	9331.1283	-448.2578	12481.9804
Industrial, commercial and transport units	1550.4505	3173.2827	2094.7855	154.4961	6973.0148
Inland waters	602.2416	267.2776	398.3284	0.0000	1267.8476
Inland wetlands	-125.9913	-96.2252	-73.2778	0.0000	-295.4943
Marine waters	-0.0002	0.0001	86.7527	0.0000	86.7526
Maritime wetlands	0.0000	0.0000	25.8566	0.0000	25.8566
Mine, dump and construction sites	-31.1788	474.9460	112.8079	-23.5198	533.0553
Open spaces with little or no vegetation	1019.9233	-287.3449	20220.0532	3005.5295	23958.1611
Pastures	-1351.7336	-788.2046	-997.1098	-110.4579	-3247.5059
Permanent crops	630.5677	13542.6842	-12524.1853	-2.3853	1646.6813
Scrub and/or herbaceous vegetation associations	-5848.5440	-5693.2856	-23714.0225	-2660.5227	-37916.3748
Urban fabric	7435.4754	4168.3497	1758.2062	5.0120	13367.0433

fig2 = plt.figure(figsize=(20, 10), dpi=100)
sub_fig = GridSpec(nrows=1, ncols=1, figure=fig2, hspace=0.35)
ax = fig2.add_subplot(sub_fig[0, 0])
lvl2_df.plot.bar(ax=ax, rot=0)

ax.yaxis.set_label_text('ha')
ax.set_xlabel('')
ax.grid(color='lightgrey', linestyle='dashdot')

plt.xticks(rotation = 90)
plt.legend(loc='best')

<matplotlib.legend.Legend at 0x7f13ab7d1fc0>

Scendendo di un livello, quindi andando ad approfondire di più il dato passando al secondo livello CLC, possiamo verificare i trend negativi a quali specifiche superfici fanno riferimento.

Le aree arbustive e con alberi radi hanno avuto la peggio, a seguire i terreni arabili, pascoli e foreste. Netto è invece l'aumento delle aree aperte, classe in cui nel terzo livello CLC ci sono le aree percorse dal fuoco(da tenere in mente per richiami successivi!). Seguono, nell'incremento, le aree urbanizzate, le aree per colture eterogenee e quelle per lo sviluppo di siti industriali e commerciali.

2.2.3 Grafici per il livello 3

lvl3 = [campania_clc1990_label3, campania_clc2000_label3, campania_clc2006_label3, campania_clc2012_label3, campania_clc2018_label3]
lvl3_df = reduce(lambda  left,right: pd.merge(left,right,on=['CORINE Land Cover - level 3'], how='outer'), lvl3)
lvl3_df.fillna(0, inplace=True)
lvl3_df.insert(loc=6, column='Surf.Diff. | 2000-1990', value=(lvl3_df.surface_ha_2000 - lvl3_df.surface_ha_1990))
lvl3_df.insert(loc=7, column='Surf.Diff. | 2006-2000', value=(lvl3_df.surface_ha_2006 - lvl3_df.surface_ha_2000))
lvl3_df.insert(loc=8, column='Surf.Diff. | 2012-2006', value=(lvl3_df.surface_ha_2012 - lvl3_df.surface_ha_2006))
lvl3_df.insert(loc=9, column='Surf.Diff. | 2018-2012', value=(lvl3_df.surface_ha_2018 - lvl3_df.surface_ha_2012))
lvl3_df.insert(loc=10, column='Surf.Diff. | 2018-1990', value=(lvl3_df.surface_ha_2018 - lvl3_df.surface_ha_1990))
lvl3_df.set_index('CORINE Land Cover - level 3', inplace=True)

lvl3_df.sort_index(ascending=True, inplace=True)
lvl3_df

	surface_ha_1990	surface_ha_2000	surface_ha_2006	surface_ha_2012	surface_ha_2018	Surf.Diff. | 2000-1990	Surf.Diff. | 2006-2000	Surf.Diff. | 2012-2006	Surf.Diff. | 2018-2012	Surf.Diff. | 2018-1990
CORINE Land Cover - level 3
Agro-forestry areas	420.5049	420.5224	140.9079	79.0823	79.0823	0.0175	-279.6145	-61.8256	0.000000e+00	-341.4226
Airports	383.8998	556.4307	556.4308	556.4308	556.4308	172.5309	0.0001	0.0000	0.000000e+00	172.5310
Annual crops associated with permanent crops	42881.8119	44016.3987	23501.2894	24459.1737	24062.2280	1134.5868	-20515.1093	957.8843	-3.969457e+02	-18819.5839
Bare rocks	1122.7257	1006.1738	979.8852	1274.5252	1274.5252	-116.5519	-26.2886	294.6400	0.000000e+00	151.7995
Beaches, dunes, sands	2282.0112	1852.6891	1835.3037	1904.4059	1904.4059	-429.3221	-17.3854	69.1022	0.000000e+00	-377.6053
Broad-leaved forest	367521.2801	369187.8047	365453.4552	367034.3194	366436.4566	1666.5246	-3734.3495	1580.8642	-5.978628e+02	-1084.8235
Burnt areas	0.0000	354.7607	438.4148	770.8451	3802.9844	354.7607	83.6541	332.4303	3.032139e+03	3802.9844
Coastal lagoons	25.8567	25.8565	25.8566	112.6093	112.6093	-0.0002	0.0001	86.7527	0.000000e+00	86.7526
Complex cultivation patterns	165047.1097	162777.2069	181971.5611	185634.2569	185582.9448	-2269.9028	19194.3542	3662.6958	-5.131210e+01	20535.8351
Coniferous forest	7457.8770	8154.5708	7620.7143	8876.9029	8373.6888	696.6938	-533.8565	1256.1886	-5.032141e+02	915.8118
Construction sites	25.4060	70.0172	246.1018	58.2792	58.2792	44.6112	176.0846	-187.8226	0.000000e+00	32.8732
Continuous urban fabric	32608.0637	33436.5752	29309.2673	33833.4142	33833.4142	828.5115	-4127.3079	4524.1469	0.000000e+00	1225.3505
Discontinuous urban fabric	37281.0655	43888.0294	52183.6870	49417.7463	49422.7583	6606.9639	8295.6576	-2765.9407	5.012000e+00	12141.6928
Dump sites	207.8811	27.2633	174.0802	438.6634	406.9530	-180.6178	146.8169	264.5832	-3.171040e+01	199.0719
Fruit trees and berry plantations	56044.2433	54999.5524	61954.0631	56585.3863	56585.3142	-1044.6909	6954.5107	-5368.6768	-7.210000e-02	541.0709
Green urban areas	731.9919	758.4238	805.2857	780.1128	1177.0586	26.4319	46.8619	-25.1729	3.969458e+02	445.0667
Industrial or commercial units	6590.0765	8168.7442	11005.1236	12643.3548	12797.8509	1578.6677	2836.3794	1638.2312	1.544961e+02	6207.7744
Inland marshes	608.8997	482.9084	386.6832	313.4054	313.4054	-125.9913	-96.2252	-73.2778	0.000000e+00	-295.4943
Land principally occupied by agriculture, with significant areas of natural vegetation	89154.5654	89342.2330	95489.3434	100261.7172	100261.7172	187.6676	6147.1104	4772.3738	0.000000e+00	11107.1518
Mineral extraction sites	1398.3376	1503.1654	1655.2099	1691.2572	1699.4478	104.8278	152.0445	36.0473	8.190600e+00	301.1102
Mixed forest	8847.1411	9446.9828	9422.1838	8444.5111	8153.6729	599.8417	-24.7990	-977.6727	-2.908382e+02	-693.4682
Moors and heathland	16.7825	0.0000	0.0000	0.0000	0.0000	-16.7825	0.0000	0.0000	0.000000e+00	-16.7825
Natural grasslands	57150.7848	53354.5398	48427.4297	30587.1997	29539.0655	-3796.2450	-4927.1101	-17840.2300	-1.048134e+03	-27611.7193
Non-irrigated arable land	304750.4005	298328.2037	283148.7731	284508.0453	284503.2765	-6422.1968	-15179.4306	1359.2722	-4.768800e+00	-20247.1240
Olive groves	58951.7577	60805.0013	66147.7874	61604.3214	61632.6846	1853.2436	5342.7861	-4543.4660	2.836320e+01	2680.9269
Pastures	15620.1214	14268.3878	13480.1832	12483.0734	12372.6155	-1351.7336	-788.2046	-997.1098	-1.104579e+02	-3247.5059
Permanently irrigated land	25316.0313	25511.2295	25683.1796	25686.6448	25718.3553	195.1982	171.9501	3.4652	3.171050e+01	402.3240
Port areas	458.6026	469.9879	481.6963	459.2550	459.2550	11.3853	11.7084	-22.4413	0.000000e+00	0.6524
Road and rail networks and associated land	650.5552	438.4218	763.6166	1242.6122	1242.6122	-212.1334	325.1948	478.9956	0.000000e+00	592.0570
Salt marshes	0.0000	0.0000	0.0000	25.8566	25.8566	0.0000	0.0000	25.8566	0.000000e+00	25.8566
Sclerophyllous vegetation	19255.7695	18537.4944	17072.9128	15623.7660	15530.6877	-718.2751	-1464.5816	-1449.1468	-9.307830e+01	-3725.0818
Sparsely vegetated areas	5561.6437	6772.6803	6445.3553	25969.2360	25942.6262	1211.0366	-327.3250	19523.8807	-2.660980e+01	20380.9825
Sport and leisure facilities	314.0751	618.1799	566.9067	650.6384	650.6384	304.1048	-51.2732	83.7317	0.000000e+00	336.5633
Transitional woodland-shrub	53654.6986	52337.4572	53035.8633	48611.2176	47091.9074	-1317.2414	698.4061	-4424.6457	-1.519310e+03	-6562.7912
Vineyards	2854.4842	2676.4992	3921.8866	1309.8441	1279.1677	-177.9850	1245.3874	-2612.0425	-3.067640e+01	-1575.3165
Water bodies	1095.5920	1373.7744	1687.8687	1981.1362	1981.1362	278.1824	314.0943	293.2675	2.273737e-13	885.5442
Water courses	472.2082	796.2674	749.4507	854.5116	854.5116	324.0592	-46.8167	105.0609	0.000000e+00	382.3034

fig3 = plt.figure(figsize=(20, 10), dpi=100)
sub_fig = GridSpec(nrows=1, ncols=1, figure=fig3, hspace=0.35)
ax = fig3.add_subplot(sub_fig[0, 0])
lvl3_df['Surf.Diff. | 2018-1990'].plot.bar(ax=ax, rot=0)

ax.yaxis.set_label_text('ha')
ax.set_xlabel('')
ax.grid(color='lightgrey', linestyle='dashdot')

plt.xticks(rotation = 90)
plt.legend(loc='best')

<matplotlib.legend.Legend at 0x7f13b28b4fd0>

Passando al livello tre della CLC, ho voluto graficare la sola differenza tra 1990 e 2018 per motivi di leggibilità del grafico. In termini di aumento di copertura del suolo c'è l'avanzamento, quasi identico, delle piccole aree destinate a colture annuali e delle aree con vegetazione frammentata. A seguire ci sono le aree urbane discontinue e quelle occupate da orti, vigneti e alberi da frutto.

Una riflessione che mi sento di fare è che l'avanzamento del tessuto urbano discontinuo ha frazionato il territorio agricolo, generando aree a piccole colture agricole.

Significativo anche l'aumento di estensione delle aree percorse dal fuoco.

I decrementi sono la cartina di tornasole degli incrementi che ho descritto poco fa e che potrebbero confermare la mia riflessione sul frazionamento del territorio naturale ed agricolo, portato avanti dall'avanzamento delle classi legati ai suoli artificiali.

2.3 Focus sui Comuni

L'analisi che segue tenta di rispondere alla domanda: "quali sono i Comuni campani in cui i fenomeni esposti poco fa hanno avuto un maggior impatto?"

Mi concentrerò sul terzo livello della CLC, in particolare sulle aree artificiali, aree agricole e foreste ma anche sulle discariche, le aree percorse dal fuoco ed il verde urbano.

Visto che dal 2000 in poi l'accuratezza geometrica del dato CLC è aumentata considerevolmente, le analisi di dettaglio che seguono si concentreranno sulla differenza tra 2000 e 2018.

# Dati 1990
focus_1990 = clc1990_campania[['pro_com', 'comune', 'provincia', 'sup_com_ha', 'label3', 'surface_ha_1990']]
focus_1990.set_index('label3', inplace=True)
focus_1990_group = pd.DataFrame(focus_1990.groupby(['pro_com', 'comune', 'sup_com_ha', 'label3'])['surface_ha_1990'].sum())
focus_1990_group.reset_index(inplace=True)

# Dati 2000
focus_2000 = clc2000_campania[['pro_com', 'comune', 'provincia', 'sup_com_ha', 'label3', 'surface_ha_2000']]
focus_2000.set_index('label3', inplace=True)
focus_2000_group = pd.DataFrame(focus_2000.groupby(['pro_com', 'comune', 'sup_com_ha', 'label3'])['surface_ha_2000'].sum())
focus_2000_group.reset_index(inplace=True)

# Dati 2006
focus_2006 = clc2006_campania[['pro_com', 'comune', 'provincia', 'sup_com_ha', 'label3', 'surface_ha_2006']]
focus_2006.set_index('label3', inplace=True)
focus_2006_group = pd.DataFrame(focus_2006.groupby(['pro_com', 'comune', 'sup_com_ha', 'label3'])['surface_ha_2006'].sum())
focus_2006_group.reset_index(inplace=True)

# Dati 2012
focus_2012 = clc2012_campania[['pro_com', 'comune', 'provincia', 'sup_com_ha', 'label3', 'surface_ha_2012']]
focus_2012.set_index('label3', inplace=True)
focus_2012_group = pd.DataFrame(focus_2012.groupby(['pro_com', 'comune', 'sup_com_ha', 'label3'])['surface_ha_2012'].sum())
focus_2012_group.reset_index(inplace=True)

# Dati 2018
focus_2018 = clc2018_campania[['pro_com', 'comune', 'provincia', 'sup_com_ha', 'label3', 'surface_ha_2018']]
focus_2018.set_index('label3', inplace=True)
focus_2018_group = pd.DataFrame(focus_2018.groupby(['pro_com', 'comune', 'sup_com_ha', 'label3'])['surface_ha_2018'].sum())
focus_2018_group.reset_index(inplace=True)

focus_1990_group

	pro_com	comune	sup_com_ha	label3	surface_ha_1990
0	61001.0	Ailano	1605.560469	Broad-leaved forest	472.5747
1	61001.0	Ailano	1605.560469	Complex cultivation patterns	93.2780
2	61001.0	Ailano	1605.560469	Continuous urban fabric	31.0972
3	61001.0	Ailano	1605.560469	Land principally occupied by agriculture, with...	188.9317
4	61001.0	Ailano	1605.560469	Non-irrigated arable land	752.7540
...	...	...	...	...	...
4543	65158.0	Bellizzi	802.200746	Continuous urban fabric	93.9079
4544	65158.0	Bellizzi	802.200746	Discontinuous urban fabric	40.9309
4545	65158.0	Bellizzi	802.200746	Fruit trees and berry plantations	24.3220
4546	65158.0	Bellizzi	802.200746	Non-irrigated arable land	426.7024
4547	65158.0	Bellizzi	802.200746	Vineyards	33.8787

4548 rows × 5 columns

2.3.1 Aree artificiali

I gruppi 1.1 ed 1.2 della CLC

# Dati 1990
artificial_surface_1990 = focus_1990_group[focus_1990_group['label3'].isin(['Continuous urban fabric', 'Discontinuous urban fabric', 'Industrial or commercial units', 'Road and rail networks and associated land', 'Port areas', 'Airports'])]
artificial_surface_1990_group = pd.DataFrame(artificial_surface_1990.groupby(['pro_com'])['surface_ha_1990'].sum())
artificial_surface_1990_group.reset_index(inplace=True)

# Dati 2000
artificial_surface_2000 = focus_2000_group[focus_2000_group['label3'].isin(['Continuous urban fabric', 'Discontinuous urban fabric', 'Industrial or commercial units', 'Road and rail networks and associated land', 'Port areas', 'Airports'])]
artificial_surface_2000_group = pd.DataFrame(artificial_surface_2000.groupby(['pro_com'])['surface_ha_2000'].sum())
artificial_surface_2000_group.reset_index(inplace=True)

# Dati 2006
artificial_surface_2006 = focus_2006_group[focus_2006_group['label3'].isin(['Continuous urban fabric', 'Discontinuous urban fabric', 'Industrial or commercial units', 'Road and rail networks and associated land', 'Port areas', 'Airports'])]
artificial_surface_2006_group = pd.DataFrame(artificial_surface_2006.groupby(['pro_com'])['surface_ha_2006'].sum())
artificial_surface_2006_group.reset_index(inplace=True)

# Dati 2012
artificial_surface_2012 = focus_2012_group[focus_2012_group['label3'].isin(['Continuous urban fabric', 'Discontinuous urban fabric', 'Industrial or commercial units', 'Road and rail networks and associated land', 'Port areas', 'Airports'])]
artificial_surface_2012_group = pd.DataFrame(artificial_surface_2012.groupby(['pro_com'])['surface_ha_2012'].sum())
artificial_surface_2012_group.reset_index(inplace=True)

# Dati 2018
artificial_surface_2018 = focus_2018_group[focus_2018_group['label3'].isin(['Continuous urban fabric', 'Discontinuous urban fabric', 'Industrial or commercial units', 'Road and rail networks and associated land', 'Port areas', 'Airports'])]
artificial_surface_2018_group = pd.DataFrame(artificial_surface_2018.groupby(['pro_com', 'comune', 'sup_com_ha'])['surface_ha_2018'].sum())
artificial_surface_2018_group.reset_index(inplace=True)

# Dati aggregati
focus_artificial = [artificial_surface_2018_group, artificial_surface_2012_group, artificial_surface_2006_group, artificial_surface_2000_group, artificial_surface_1990_group]
focus_artificial_df = reduce(lambda  left,right: pd.merge(left,right,on=['pro_com'], how='outer'), focus_artificial)
focus_artificial_df = focus_artificial_df[focus_artificial_df['sup_com_ha'] > 0]
focus_artificial_df.fillna(0, inplace=True)
focus_artificial_df.insert(loc=3, column='perc_1990', value=round(focus_artificial_df.surface_ha_1990 / focus_artificial_df.sup_com_ha, 4) * 100)
focus_artificial_df.insert(loc=4, column='perc_2000', value=round(focus_artificial_df.surface_ha_2000 / focus_artificial_df.sup_com_ha, 4) * 100)
focus_artificial_df.insert(loc=5, column='perc_2006', value=round(focus_artificial_df.surface_ha_2006 / focus_artificial_df.sup_com_ha, 4) * 100)
focus_artificial_df.insert(loc=6, column='perc_2012', value=round(focus_artificial_df.surface_ha_2012 / focus_artificial_df.sup_com_ha, 4) * 100)
focus_artificial_df.insert(loc=7, column='perc_2018', value=round(focus_artificial_df.surface_ha_2018 / focus_artificial_df.sup_com_ha, 4) * 100)
focus_artificial_df.insert(loc=8, column='Perc.Diff. | 2018-2000', value=(focus_artificial_df.perc_2018 - focus_artificial_df.perc_2000))
focus_artificial_df.sort_values(by=['Perc.Diff. | 2018-2000'], ascending=True, inplace=True)
focus_artificial_df

	pro_com	comune	sup_com_ha	perc_1990	perc_2000	perc_2006	perc_2012	perc_2018	Perc.Diff. | 2018-2000	surface_ha_2018	surface_ha_2012	surface_ha_2006	surface_ha_2000	surface_ha_1990
201	63036.0	Grumo Nevano	287.688759	47.45	64.93	64.93	51.31	51.31	-13.62	147.6065	147.6065	186.8019	186.8019	136.5066
232	63067.0	San Giorgio a Cremano	415.383936	60.96	76.57	68.51	68.51	68.51	-8.06	284.5767	284.5767	284.5768	318.0615	253.2174
197	63032.0	Frattamaggiore	538.816178	58.54	71.14	72.46	63.97	63.97	-7.17	344.6705	344.6705	390.4469	383.2980	315.4185
476	65121.0	San Mango Piemonte	602.370147	7.81	7.81	3.14	3.14	3.14	-4.67	18.8890	18.8890	18.8890	47.0198	47.0222
137	62046.0	Pago Veiano	2374.895020	1.83	5.24	2.13	2.13	2.13	-3.11	50.6037	50.6037	50.6037	124.5097	43.5550
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
228	63063.0	Quarto	1415.896056	24.40	24.95	26.90	39.26	39.26	14.31	555.9065	555.9065	380.8373	353.2075	345.4612
254	63089.0	Volla	620.630205	45.68	49.86	49.94	64.34	64.34	14.48	399.3008	399.3008	309.9351	309.4482	283.5034
212	63047.0	Monte di Procida	370.428037	31.25	31.25	31.25	48.28	48.28	17.03	178.8412	178.8412	115.7459	115.7458	115.7439
190	63025.0	Castello di Cisterna	392.135970	21.89	46.04	68.31	68.31	68.31	22.27	267.8514	267.8514	267.8514	180.5292	85.8311
380	65011.0	Atrani	12.061182	0.00	0.00	52.01	52.01	52.01	52.01	6.2730	6.2730	6.2730	0.0000	0.0000

507 rows × 14 columns

# Superfici in arretramento
reduction_artificial_area = focus_artificial_df.iloc[0:10]
reduction_artificial_area

	pro_com	comune	sup_com_ha	perc_1990	perc_2000	perc_2006	perc_2012	perc_2018	Perc.Diff. | 2018-2000	surface_ha_2018	surface_ha_2012	surface_ha_2006	surface_ha_2000	surface_ha_1990
201	63036.0	Grumo Nevano	287.688759	47.45	64.93	64.93	51.31	51.31	-13.62	147.6065	147.6065	186.8019	186.8019	136.5066
232	63067.0	San Giorgio a Cremano	415.383936	60.96	76.57	68.51	68.51	68.51	-8.06	284.5767	284.5767	284.5768	318.0615	253.2174
197	63032.0	Frattamaggiore	538.816178	58.54	71.14	72.46	63.97	63.97	-7.17	344.6705	344.6705	390.4469	383.2980	315.4185
476	65121.0	San Mango Piemonte	602.370147	7.81	7.81	3.14	3.14	3.14	-4.67	18.8890	18.8890	18.8890	47.0198	47.0222
137	62046.0	Pago Veiano	2374.895020	1.83	5.24	2.13	2.13	2.13	-3.11	50.6037	50.6037	50.6037	124.5097	43.5550
485	65132.0	San Valentino Torio	915.632363	15.51	15.56	13.27	13.27	13.27	-2.29	121.4587	121.4587	121.4588	142.4372	141.9949
4	61005.0	Aversa	885.176217	59.33	67.87	67.96	66.24	66.24	-1.63	586.3294	586.3294	601.5394	600.7846	525.1697
170	63005.0	Arzano	473.103075	83.22	86.97	87.40	85.42	85.42	-1.55	404.1348	404.1348	413.4904	411.4785	393.7116
474	65119.0	San Giovanni a Piro	3790.148485	2.12	4.21	2.72	2.72	2.72	-1.49	103.0783	103.0783	103.0783	159.6782	80.3555
57	61067.0	Recale	321.954739	32.81	32.81	31.54	31.54	31.54	-1.27	101.5467	101.5467	101.5467	105.6199	105.6193

# Superfici in avanzamento
increase_artificial_area = focus_artificial_df.iloc[-10:]
increase_artificial_area.sort_values(by=['Perc.Diff. | 2018-2000'], ascending=False, inplace=True)
increase_artificial_area

	pro_com	comune	sup_com_ha	perc_1990	perc_2000	perc_2006	perc_2012	perc_2018	Perc.Diff. | 2018-2000	surface_ha_2018	surface_ha_2012	surface_ha_2006	surface_ha_2000	surface_ha_1990
380	65011.0	Atrani	12.061182	0.00	0.00	52.01	52.01	52.01	52.01	6.2730	6.2730	6.2730	0.0000	0.0000
190	63025.0	Castello di Cisterna	392.135970	21.89	46.04	68.31	68.31	68.31	22.27	267.8514	267.8514	267.8514	180.5292	85.8311
212	63047.0	Monte di Procida	370.428037	31.25	31.25	31.25	48.28	48.28	17.03	178.8412	178.8412	115.7459	115.7458	115.7439
254	63089.0	Volla	620.630205	45.68	49.86	49.94	64.34	64.34	14.48	399.3008	399.3008	309.9351	309.4482	283.5034
228	63063.0	Quarto	1415.896056	24.40	24.95	26.90	39.26	39.26	14.31	555.9065	555.9065	380.8373	353.2075	345.4612
82	61094.0	Trentola Ducenta	665.513019	30.18	30.19	31.44	44.05	44.05	13.86	293.1760	293.1760	209.2403	200.8859	200.8848
15	61016.0	Carinaro	631.593670	28.55	34.93	44.25	47.41	47.41	12.48	299.4077	299.4077	279.4878	220.6269	180.3185
80	61092.0	Teverola	669.906081	31.59	37.22	50.70	48.78	48.78	11.56	326.7535	326.7535	339.6324	249.3457	211.6063
188	63023.0	Casoria	1213.840286	57.66	72.14	72.14	83.48	83.48	11.34	1013.2593	1013.2593	875.7210	875.7214	699.8511
123	62032.0	Forchia	545.454120	0.00	0.00	0.00	11.00	11.00	11.00	59.9761	59.9761	0.0000	0.0000	0.0000

artificial_df = pd.concat([reduction_artificial_area, increase_artificial_area])
artificial_df.sort_values(by=['Perc.Diff. | 2018-2000'], ascending=True, inplace=True)
artificial_df = artificial_df[['comune', 'Perc.Diff. | 2018-2000']]
artificial_df['Perc.Diff. | 2018-2000'] = artificial_df['Perc.Diff. | 2018-2000']
artificial_df.set_index('comune', inplace=True)

fig_artificial = plt.figure(figsize=(20, 10), dpi=100)
sub_fig_artificial = GridSpec(nrows=1, ncols=1, figure=fig_artificial, hspace=0.35)
ax_artificial = fig_artificial.add_subplot(sub_fig_artificial[0, 0])
artificial_df.plot.bar(ax=ax_artificial, rot=0)

ax_artificial.yaxis.set_label_text('%')
ax_artificial.set_xlabel('')
ax_artificial.grid(color='lightgrey', linestyle='dashdot')

plt.xticks(rotation = 90)
plt.legend(loc='best')

<matplotlib.legend.Legend at 0x7f13b50e6e60>

Onestamente non so come interpretare l'arretramento delle aree artificiali, andrebbe fatta una ulteriore analisi per valutare se sono Comuni che si sono spopolati in quasi trenta anni.

L'aumento è abbastanza facile da indagare per me che sono del posto. Napoli ha una popolazione di quasi un milione di abitanti che triplica nei giorni feriali, per non parlare poi dei turisti. E' un'area molto congestionata con possibilità di nuova edilizia prossime allo zero; risulta normale quindi che Comuni più prossimi al Capoluogo di Regione stiano accogliendo "l'esubero" di popolazione. Faccio riferimento a Castello di Cisterna, Volla e Quarto.

Il caso di Atrani è particolare perchè è il Comune più piccolo d'Italia ed è posto in una zona abbastanza impervia della costiera amalfitana; è probabile che nel 2000 non si sia data molta importanza all'area.

2.3.2 Aree forestate

Il gruppo 3.1 della CLC

# Dati 1990
forest_surface_1990 = focus_1990_group[focus_1990_group['label3'].isin(['Broad-leaved forest', 'Coniferous forest', 'Mixed forest'])]
forest_surface_1990_group = pd.DataFrame(forest_surface_1990.groupby(['pro_com'])['surface_ha_1990'].sum())
forest_surface_1990_group.reset_index(inplace=True)

# Dati 2000
forest_surface_2000 = focus_2000_group[focus_2000_group['label3'].isin(['Broad-leaved forest', 'Coniferous forest', 'Mixed forest'])]
forest_surface_2000_group = pd.DataFrame(forest_surface_2000.groupby(['pro_com'])['surface_ha_2000'].sum())
forest_surface_2000_group.reset_index(inplace=True)

# Dati 2006
forest_surface_2006 = focus_2006_group[focus_2006_group['label3'].isin(['Broad-leaved forest', 'Coniferous forest', 'Mixed forest'])]
forest_surface_2006_group = pd.DataFrame(forest_surface_2006.groupby(['pro_com'])['surface_ha_2006'].sum())
forest_surface_2006_group.reset_index(inplace=True)

# Dati 2012
forest_surface_2012 = focus_2012_group[focus_2012_group['label3'].isin(['Broad-leaved forest', 'Coniferous forest', 'Mixed forest'])]
forest_surface_2012_group = pd.DataFrame(forest_surface_2012.groupby(['pro_com'])['surface_ha_2012'].sum())
forest_surface_2012_group.reset_index(inplace=True)

# Dati 2018
forest_surface_2018 = focus_2018_group[focus_2018_group['label3'].isin(['Broad-leaved forest', 'Coniferous forest', 'Mixed forest'])]
forest_surface_2018_group = pd.DataFrame(forest_surface_2018.groupby(['pro_com', 'comune', 'sup_com_ha'])['surface_ha_2018'].sum())
forest_surface_2018_group.reset_index(inplace=True)

# Dati aggregati
focus_forest = [forest_surface_2018_group, forest_surface_2012_group, forest_surface_2006_group, forest_surface_2000_group, forest_surface_1990_group]
focus_forest_df = reduce(lambda  left,right: pd.merge(left,right,on=['pro_com'], how='outer'), focus_forest)
focus_forest_df = focus_forest_df[focus_forest_df['sup_com_ha'] > 0]
focus_forest_df.fillna(0, inplace=True)
focus_forest_df.insert(loc=3, column='perc_1990', value=round(focus_forest_df.surface_ha_1990 / focus_forest_df.sup_com_ha, 4) * 100)
focus_forest_df.insert(loc=4, column='perc_2000', value=round(focus_forest_df.surface_ha_2000 / focus_forest_df.sup_com_ha, 4) * 100)
focus_forest_df.insert(loc=5, column='perc_2006', value=round(focus_forest_df.surface_ha_2006 / focus_forest_df.sup_com_ha, 4) * 100)
focus_forest_df.insert(loc=6, column='perc_2012', value=round(focus_forest_df.surface_ha_2012 / focus_forest_df.sup_com_ha, 4) * 100)
focus_forest_df.insert(loc=7, column='perc_2018', value=round(focus_forest_df.surface_ha_2018 / focus_forest_df.sup_com_ha, 4) * 100)
focus_forest_df.insert(loc=8, column='Perc.Diff. | 2018-2000', value=(focus_forest_df.perc_2018 - focus_forest_df.perc_2000))
focus_forest_df.sort_values(by=['Perc.Diff. | 2018-2000'], ascending=True, inplace=True)
focus_forest_df

	pro_com	comune	sup_com_ha	perc_1990	perc_2000	perc_2006	perc_2012	perc_2018	Perc.Diff. | 2018-2000	surface_ha_2018	surface_ha_2012	surface_ha_2006	surface_ha_2000	surface_ha_1990
139	62073.0	Solopaca	3112.664003	43.63	43.63	43.46	11.69	11.69	-31.94	363.9901	363.9901	1352.7210	1358.0359	1358.0275
143	62077.0	Vitulano	3599.050416	50.86	50.81	50.81	26.82	26.82	-23.99	965.1276	965.1276	1828.5273	1828.5272	1830.3353
410	65115.0	Salento	2378.507005	51.73	50.93	37.94	37.93	37.93	-13.00	902.2741	902.2741	902.4736	1211.3768	1230.4272
179	63082.0	Terzigno	2346.480829	20.68	20.68	20.68	20.68	8.01	-12.67	187.8420	485.2334	485.2333	485.2338	485.2356
147	63007.0	Barano d'Ischia	1096.089426	49.21	49.21	49.21	38.59	38.59	-10.62	422.9915	422.9915	539.3662	539.3660	539.3685
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
439	65147.0	Torchiara	846.167482	20.76	20.76	20.86	29.14	29.14	8.38	246.5860	246.5860	176.5072	175.6504	175.6495
436	65144.0	Stella Cilento	1451.726154	39.51	43.25	43.25	52.64	52.64	9.39	764.2043	764.2043	627.8681	627.8690	573.5525
393	65098.0	Pollica	2816.525386	11.44	17.14	17.15	27.01	27.01	9.87	760.6785	760.6785	482.9555	482.8660	322.2088
359	65064.0	Lustra	1524.153297	34.01	34.02	34.02	44.61	44.61	10.59	679.9189	679.9189	518.4959	518.4971	518.4368
408	65113.0	Sacco	2365.702940	42.11	42.02	42.02	63.01	63.01	20.99	1490.6391	1490.6391	994.0414	994.0408	996.2456

450 rows × 14 columns

# Superfici in arretramento
reduction_forest_area = focus_forest_df.iloc[0:10]
reduction_forest_area

	pro_com	comune	sup_com_ha	perc_1990	perc_2000	perc_2006	perc_2012	perc_2018	Perc.Diff. | 2018-2000	surface_ha_2018	surface_ha_2012	surface_ha_2006	surface_ha_2000	surface_ha_1990
139	62073.0	Solopaca	3112.664003	43.63	43.63	43.46	11.69	11.69	-31.94	363.9901	363.9901	1352.7210	1358.0359	1358.0275
143	62077.0	Vitulano	3599.050416	50.86	50.81	50.81	26.82	26.82	-23.99	965.1276	965.1276	1828.5273	1828.5272	1830.3353
410	65115.0	Salento	2378.507005	51.73	50.93	37.94	37.93	37.93	-13.00	902.2741	902.2741	902.4736	1211.3768	1230.4272
179	63082.0	Terzigno	2346.480829	20.68	20.68	20.68	20.68	8.01	-12.67	187.8420	485.2334	485.2333	485.2338	485.2356
147	63007.0	Barano d'Ischia	1096.089426	49.21	49.21	49.21	38.59	38.59	-10.62	422.9915	422.9915	539.3662	539.3660	539.3685
304	65006.0	Amalfi	570.014402	48.51	48.51	38.86	38.86	38.86	-9.65	221.4838	221.4838	221.4838	276.5111	276.5144
170	63064.0	Ercolano	1989.359843	21.47	21.26	21.26	17.39	11.75	-9.51	233.6730	345.9226	422.9398	422.9397	427.1795
155	63035.0	Gragnano	1464.244074	53.35	53.35	52.42	44.07	44.07	-9.28	645.3063	645.3063	767.6083	781.1773	781.1751
159	63041.0	Marano di Napoli	1565.253641	13.09	13.09	13.09	3.98	3.98	-9.11	62.3734	62.3734	204.9146	204.9145	204.9150
183	63091.0	Trecase	621.312671	8.86	8.86	8.86	8.86	0.08	-8.78	0.4777	55.0368	55.0368	55.0368	55.0371

# Superfici in avanzamento
increase_forest_area = focus_forest_df.iloc[-10:]
increase_forest_area.sort_values(by=['Perc.Diff. | 2018-2000'], ascending=False, inplace=True)
increase_forest_area

	pro_com	comune	sup_com_ha	perc_1990	perc_2000	perc_2006	perc_2012	perc_2018	Perc.Diff. | 2018-2000	surface_ha_2018	surface_ha_2012	surface_ha_2006	surface_ha_2000	surface_ha_1990
408	65113.0	Sacco	2365.702940	42.11	42.02	42.02	63.01	63.01	20.99	1490.6391	1490.6391	994.0414	994.0408	996.2456
359	65064.0	Lustra	1524.153297	34.01	34.02	34.02	44.61	44.61	10.59	679.9189	679.9189	518.4959	518.4971	518.4368
393	65098.0	Pollica	2816.525386	11.44	17.14	17.15	27.01	27.01	9.87	760.6785	760.6785	482.9555	482.8660	322.2088
436	65144.0	Stella Cilento	1451.726154	39.51	43.25	43.25	52.64	52.64	9.39	764.2043	764.2043	627.8681	627.8690	573.5525
439	65147.0	Torchiara	846.167482	20.76	20.76	20.86	29.14	29.14	8.38	246.5860	246.5860	176.5072	175.6504	175.6495
405	65110.0	Romagnano al Monte	966.995480	18.35	18.35	18.35	26.64	26.64	8.29	257.5984	257.5984	177.4251	177.4254	177.4249
151	63022.0	Casola di Napoli	259.318212	34.28	34.28	42.15	42.15	42.15	7.87	109.3078	109.3078	109.3078	88.8917	88.8916
102	62036.0	Ginestra degli Schiavoni	1478.706162	22.20	22.20	22.20	29.90	29.90	7.70	442.1457	442.1457	328.2494	328.2500	328.2562
229	64045.0	Luogosano	606.809422	10.43	12.63	12.63	18.42	18.42	5.79	111.7932	111.7932	76.6287	76.6287	63.2980
355	65060.0	Laureana Cilento	1374.420460	25.62	25.64	25.64	30.10	31.15	5.51	428.1047	413.7489	352.3810	352.3819	352.1644

forest_df = pd.concat([reduction_forest_area, increase_forest_area])
forest_df.sort_values(by=['Perc.Diff. | 2018-2000'], ascending=True, inplace=True)
forest_df = forest_df[['comune', 'Perc.Diff. | 2018-2000']]
forest_df['Perc.Diff. | 2018-2000'] = forest_df['Perc.Diff. | 2018-2000']
forest_df.set_index('comune', inplace=True)

fig_forest = plt.figure(figsize=(20, 10), dpi=100)
sub_fig_forest = GridSpec(nrows=1, ncols=1, figure=fig_forest, hspace=0.35)
ax_forest = fig_forest.add_subplot(sub_fig_forest[0, 0])
forest_df.plot.bar(ax=ax_forest, rot=0)

ax_forest.yaxis.set_label_text('%')
ax_forest.set_xlabel('')
ax_forest.grid(color='lightgrey', linestyle='dashdot')

plt.xticks(rotation = 90)
plt.legend(loc='best')

<matplotlib.legend.Legend at 0x7f13b1292cb0>