Results

The Result

1. Land use data of Mae Huad Sector, Ngo Demonstration Forest

Land use had been attempted to classfified into 5 types as following (Table 1 and Figure 1&2);

1.1 Urban and built-up area;

Urban and built up area (U) consist of all residential area, village, institute, temple, school and etc. It cover an area of 3.3572 square kilometers or 0.734% of the total area. It locate many ancient settlements namely; Ban (or village) On, Ban Lai, Ban Pong Ma O, Ban Pong Tong, Ban Mai, Ban Sop Plung, Ban Had Cheo, Ban Rong Tar, Ban Pong, Ban Tak, Ban Huad, Ban Suan Sak, Ban Pang Lah and Ban Prow, and  most of them are lowland local people settlements. While Ban Mae Wit, Ban Sop Not, Ban Huai Not, Ban Huai Hok,, Ban Maeo Mae Klang, Ban Huai Nam Rin, Ban Maeo Mae Ka and Ban Kun Mae Huad are mong villages that all settle on the upper part of montane area.

            1.2 Agricultural Land (divided into four sub-categories) that are;

Agricultural land contain various kinds of both annual cash crops and perennial crops. The most vast area is field crops land (FC) or 29.3448 square kilometers (or 6.412 % of the total). Owners grow different kinds of cash crops, for example; maiz for livestock, sugar cane, cassava, pine apple, vegetable garden and others. These field crops usually alter due to growing season, site, soil, water resource and topography. It also include horticultural crops or fruit tree orchards such as; lichii,  longan, tamarine, orange, banana and etc. 

            The paddy land (PD), almost are rainfed rice field when the farmer grow rice once in a year during rainy season. After harvesting, they may grow some onion, garlic, chili and other vegetable for household consumption.  Paddy land occupied the area of 8.5226 square kilometers or 1.862 % of the total area. It is found on flat area where close to the village and along the river levee, few are distribute on high land.

            Field crops (FC), imply for many kinds of cash crops, i.e., maize, sugar cane, cassava, pine apple, peanut, tobacco, chili, vegetable garden and others. Selection of crops is un-certain due to market price or demand and supply, then activity in these field area are also varied each year. Some year, the land are left. It found that 29.3448 square kilometers or 6.412 % of the total area. Their location are next to paddy land, close to the wood land. More rigid situation, when the interpretation show the vast area of field crops expand into the forest land on the upper part of Mae Huad watershed area, where are important for gathering the water to lower down steam, especially on the west and south parts of the project.

            Fruit orchards (PO) comprise of various fruit tree orchards such as; lichii,  longan, tamarine, orange, banana and etc. The good site for orchard trees are along hill slope until highland. It found that 1.9671 square kilometers or 0.430 % of the total area are cover by fruit trees.

            Miscellaneous land (MS) imply for others in agriculture category, when farmer leave their land for some years, then many grass and weed are occupy rapidly. But they will repeat their activities at sometimes. The other mean miscellaneous crops, bamboo garden, small vegetable garden for household usage, tree nursery. It occupy on flat to undulating area or alternate to field crop land. Miscellaneous land cover an area of 6.3066 square kilometers or 1.378 % of the total area.

 

            1.3 Forest Land (classified into four sub forest eco-types) that are;

            Evergreen forest (EV) are rarely found along summit of the west side of the project where are the topographic dividing ridge between (Amphoe) Ngao and Chae Hom districts. They are destroyed by local people and hill-tribe man. We found many evergreen forest has transformed into shifting cultivation or slash and burn area. The site of evergreen forest are found in small pocket in the valley and some are narrow strip along the stream. This sub-type occupy the area of 3.2333 square kilometers or 0.706 % of the total area.

            Mixed deciduous forest (MD) are found to be the most vast forest area of study site. Deciduous tree are almost shed their leaves during dry period annually. About 358.1564 square kilometers out of  457.6788 square kilometers or 78.255 % of the total area cover by mixed deciduous forest. This sub-eco-type occupied a wide range of elevation where ranging between 300 and up to 900 meters above mean sea level, but usually found on undulating terrain. Distinguish phenomena, is comprising of teak in their stand composition where limestone are soil parent material. Forest fire are occur during dry period annually.

            Dry deciduous dipterocarp (DD) are forest eco-type when they are rich in deciduous tree species all of them shed their leaves when are in drought period, along with forest fire, that why they appear dark on Landsat image and become red again in rainy season. This type are similar in shedding as mixed deciduous, but are different in tree species composition, those are more drought tolerant. We usually found them along the ridge of east part. Dry dipterocarp forest also found on rough terrain with shallow soil liked the mining valley. It occupy the area of 18.2481 square kilometers or 3.987 % of the total area.

            Disturbed forest (DF) is severe type of forest when all trees had been cut in order to alter the forest to agricultural land. When they slash and burn a large area and left abandoned, they will be success by light demanding trees (non-economic species), grass and weed. The classification show the area of 13.6881 square kilometers or 2.991 % of the total area are deteriorate forest.

            Forest plantation (PT) are aimed for economic and gene pool conservation of teak. Teak plantation for conservative purpose had been done since 1942 or the oldest teak are now 76 years. To productive purpose, the government had initiate many projects concerned with teak forest plantation promotion. The projects have providing both fund and teak seedling to local people to increase forest area and commercial teak for lumber. Very few are other species or para-rubber tree that introduce since 2010. Many of teak plantations have invaded and occupied by local people, many good quality teak are illegal cut. The area of plantation are 12.8523 square kilometers or 2.808 % of the total area.

Table 1 Land use classification of  Mae Huad Sec tor, Ngao Demonstration Forest, Lampang province

Notation : Base on visual interpretation of Color ortho-aerial photographs (1:4,00 scale) taken

     In 2002 and up-to-dated by Landsat-8 imageries acquired during 2016 to the late of 2017.

Figure 1 Land use categories of  Mae Huad Sector, Ngao Demonstration Forest, Lampang province illustrated in square kilometers and percentage

1.4 Water bodies

Water bodies are included both natural and man-made ponds (W). They are small in size, not bigger than half an acre. They situate in rice field and farm land for agricultural purpose.  Surface water area are estimate at 0.2669 square kilometers or 0.058 % of the total area. Other water resource or river are also delineated as poly line based on 1:50,000 topographic maps, and compile separate layer.

 

1.5 Miscellaneous use of land or others

It imply for two facilities. Transmission line (MS1) those belong to the Electric Generation Authority of Thailand. The line lay north to south and connect between two provinces; Phayao and Lampang. They clear all trees under the line and lost forest land when they pass through forest with 50 meters width. It is estimated at 0.8753 square kilometers or 0.191% of the total area. The other miscellaneous use of land is highway number 2, which connect transportation between Phayao and Lampang. The paved road is estimate at 0.8602 square kilometers or 0.188 % of the total area.

Figure 1 Forest Land use classification map of Mae Huad Project area

Notation: DD = Dry Deciduous        DF  = Disturbed Forest           EV = Evergreen Forest

                 FC  = Field Crops               MD = Mixed Deciduous          MS = Miscellaneous

                 MS1= Transmission Line    MS2 = Pave Road                    PD = Paddy Land

                 PO = Fruit Orchard            PT   = Teak Plantation              U  = Urban Land 

                 W   = Water Bosies

 

2. Field inventory for species list

The Summarized field data to obtain per-hectare plot statistics, including basal area, number of species, wood density by specie and Importance Value Index (IVI) was shown below;

• 54 sampling points in 3 forest types with average BA of 18.39 m²/ha and 549 tree/ha.
• 46, 18 and 32 tree species were found in MDF, DDF and DEF, respectively.
• Tree species with the highest IVI in MDF are Xylia xylocarpa, Tectona grandis and Pterocarpus macrocarpus
• Tree species with the highest IVI in DDF are Shorea siamensis, obtusa and Pterocarpus macrocarpus
• Tree species with the highest IVI in DEF are Croton Roxburghii, Hopea odorata and Duabanga grandiflora

 

3. Tree species classification

Tree wood density obtained from published data were used as a guidelines to classify tree species in each forest type into 10 groups (30 groups in total). In each group, the major tree species with the highest IVI was selected for sample collection (Table 1-3).

 

Table 1 The selected species for sample collection in the Mixed Deciduous Forest

Class No.	Range of wood density (kg/m2)	Major species (Scientific name)
1	282-385	Cananga  latifolia.   Finet & Gagnep.
2	386-488	Litsea  glutinosa    C.B. Robinson
3	489-591	Lannea  coromandelica    Merr.
4	592-694	Tectona  grandis    Linn. f.
5	695-797	Albizia  odoratissima    Benth.
6	798-900	Terminalia  nigrovenulosa    Pierre ex Laness.
7	901-1003	Pterocarpus  macrocarpus    Kurz
8	1004-1106	Xylia  xylocarpa    Taub.
9	1107-1209	Dalbergia oliveri Gamble.
10	1210-1312	Terminalia  corticosa    Pierre ex Laness.

 

Table 2 The selected species for sample collection in the Dry Dipterocarp Forest

Class No.	Range of wood density (kg/m2)	Major species (Scientific name)
1	400-485	Mitragyna brunonis Craib
2	486-570	Bridelia pierrei Gagnep.
3	571-655	Gardenia sootepensis Hutch.
4	656-740	Haldina cordifolia (Roxb.) Ridsdale.
5	741-825	Dipterocarpus obtusifolius Teijsm. ex Miq.
6	826-910	NA
7	911-995	Pterocarpus macrocarpus Kurz
8	996-1080	Shorea siamensis Miq.
9	1081-1165	Dalbergia oliveri Gamble ex Prain.
10	1166-1250	Terminalia corticosa Pierre ex Laness.

 

Table 3 The selected species for sample collection in the Dry Evergreen Forest

Class No.	Range of wood density (kg/m2)	Major species (Scientific name)
1	387-474	Duabanga grandiflora Walp.
2	475-561	Croton roxburghii N.P.Balakr.
3	562-648	Careya sphaerica Roxb.
4	649-735	Artocarpus lakoocha Roxb.
5	736-822	Cratoxylum formosum (Jack) Dyer.
6	823-909	Anogeissus acuminata Wall.
7	910-996	Pterocarpus macrocarpus Kurz
8	997-1083	Terminalia alata Heyne ex Roth
9	1084-1170	Xylia xylocarpa Taub.
10	1171-1257	Quercus kerrii Craib

 

4. Wood Carbon Fraction Analysis

The extracted wood samples were weighted, dried, re-weighed and pulverized to analyze the carbon content in the laboratory using the C/N analyzer. Carbon contents of the selected tree species in each forest type are shown in table 1-3.

Table 1 Carbon contents of sample tree species in the Mixed Deciduous Forest

Range of wood density (kg/m2)	Major Species (Scientific name)	No. of sample trees	Cabon content (%)
282-385	Cananga  latifolia.   Finet & Gagnep.	15	47.75
386-488	Litsea  glutinosa    C.B. Robinson	15	46.86
489-591	Lannea  coromandelica    Merr.	16	45.75
592-694	Tectona  grandis    Linn. f.	16	49.66
695-797	Albizia  odoratissima    Benth.	15	46.84
798-900	Terminalia  nigrovenulosa    Pierre ex Laness.	16	47.13
901-1003	Pterocarpus  macrocarpus    Kurz	15	48.41
1004-1106	Xylia  xylocarpa    Taub.	15	48.03
1107-1209	Dalbergia oliveri Gamble.	17	47.13
1210-1312	Terminalia  corticosa    Pierre ex Laness.	15	48.55

 

Table 2 Carbon contents of sample tree species in the Dry Dipterocarp Forest

Range of wood density (kg/m2)	Major Species (Scientific name)	No. of sample trees	Cabon content (%)
400-485	Mitragyna brunonis Craib	15	47.57
486-570	Bridelia pierrei Gagnep.	12	47.16
571-655	Gardenia sootepensis Hutch.	15	46.06
656-740	Haldina cordifolia (Roxb.) Ridsdale.	15	48.262
741-825	Dipterocarpus obtusifolius Teijsm. ex Miq.	15	47.62
826-910	NA
911-995	Pterocarpus macrocarpus Kurz	15	48.41
996-1080	Shorea siamensis Miq.	15	46.76
1081-1165	Dalbergia oliveri Gamble ex Prain.	17	47.13
1166-1250	Terminalia corticosa Pierre ex Laness.	15	48.55

Table 3 Carbon contents of sample tree species in the Dry Evergreen Forest

Range of wood density (kg/m2)	Major Species (Scientific name)4.3	No. of sample trees	Cabon content (%)
387-474	Duabanga grandiflora Walp.	15	46.92
475-561	Croton roxburghii N.P.Balakr.	15	47.77
562-648	Careya sphaerica Roxb.	15	47.47
649-735	Artocarpus lakoocha Roxb.	15	48.31
736-822	Cratoxylum formosum (Jack) Dyer.	15	46.83
823-909	Anogeissus acuminata Wall.	15	46.81
910-996	Pterocarpus macrocarpus Kurz	15	48.41
997-1083	Terminalia nigrovenulosa Pierre ex Laness.	15	45.75
1084-1170	Xylia xylocarpa Taub.	15	48.03
1171-1257	Quercus kerrii Craib	15	45.43

 

5. Tree Bole Carbon Equations

5.1 Tree Bole Carbon Equations in the Mixed Deciduous Forest

Ten tree bole carbon equations derived from the Mixed Deciduous Forest were constructed based on wood density that ranged between 282-1,312 kg/m³ (Table 1). A general equation which was used for tree species in all wood density groups in the Mixed Deciduous Forest was also constructed (Eq. 1).

 

C=  0.018155 D^2.2204 H ^0.490…………………………..( Eq. 1)

where; C = Carbon storage in stem bole, kg/tree

                                    D = Diameter at breast height of tree, cm

                                    H = Total height of tree, m

Table 1 The carbon equations classified by wood density of tree species in the mixed deciduous forest

No.	Sample Species	Carbon Equations	DBH Range (cm)
1	Ficus  var.pubescens Cananga  latifolia Bombax  insulare	C = 0.008730 D2.335 H 0.570	13.2-43
2	Tetrameles nudiflora Elaeocarpus  stipularis Croton roxburghii Grewia  elastica Litsea  glutinosa Sterculia  pexa Ailanthus  triphysa	C = 0.019454 D2.335 H 0.338	16.2-63
3	Cleidion  spiciflorum Lannea  coromandelica Canarium  subulatum Miliusa  velutina	C = 0.001538 D3.014 H 0.475	11.8-58
4	Radermachera  pierrei Tectona  grandis Lagerstroemia  duperreana Terminalia nigrovenulosa	C = 0.018836 D1.833 H 0.848	8.7-71
5	Buchanania  latifolia  Spondias bipinnata  Dipterocarpus  turbinatus Dipterocarpus  costatus Albizia  odoratissima  Terminalia  bellerica  Lagerstroemia  macrocarpa  Dillenia obovata	C = 0.011350 D2.043 H 0.853	11.0-29
6	Stereospermum  neuranthum  Anogeissus  acuminata  Terminalia  nigrovenulosa  Vitex  canescens  Chukrasia  velutina  Eugenia  cumini  Vitex  peduncularis	C = 0.067764 D2.011 H 0.277	15-69
7	Pterocarpus  macrocarpus  Madhuca thorelii  Diospyros  ehretioides	C = 0.014093 D2.068 H 0.723	11.5-61.5
8	Xylia  xylocarpa  Millettia  brandisiana Irvingia  malayana  Terminalia  alata  Schleichera  oleosa	C = 0.011967 D2.067 H 0.791	13.2-68.8
9	Butea  monosperma  Dalbergia oliveri	C = 0.017539 D2.276 H 0.547	11.1-42.8
10	Quercus  kerrii  Terminalia  corticosa  Diospyros  mollis	C = 0.005957 D2.206 H 0.819	13.2-66.5

 

5.2 Tree Bole Carbon Equations in the Dry Dipterocarp Forest

Nine tree bole carbon equations derived from the Dry Dipterocarp Forest were constructed based on wood density ranged between 400-1,250 kg/m³ (Table 2). A general equation which was used for all tree species in the Dry Diptercarp Forest was also constructed (Eq. 2) as followed:

 

              C = 0.009462 D ^2.328 H ^0.602…………………………..(Eq. 2)

           

where; C = Carbon storage in stem bole, kg/tree

                                    D = Diameter at breathn height of tree, cm

                                    H = Total height of tree, m

Table 2 The carbon equations classified by wood density of tree species in the Dry Dipterocarp Forest

No.	Sample Species	Carbon Equations	DBH Range (cm)
1	Mitragyna brunonis	C = 0.006353 D2.227 H 0.802	13-44.1
2	Bridelia pierrei	C = 0.004887 D2.618 H 0.438	10-28.6
3	Gardenia sootepensis	C = 0.020417 D2.237 H 0.696	11-2.4
4	Haldina cordifolia Buchanania  latifolia	C = 0.001928 D2.664 H 0.679	10.2-41.9
5	Dipterocarpus obtusifolius.	C = 0.000975 D2.389 H 1.277	13.1-42.5
6	N/A
7	Dalbergia assamica Pterocarpus macrocarpus	C = 0.014093 D2.068 H 0.723	11.5-61.5
8	Shorea siamensis  Millettia brandisiana  Shorea obtusa  Terminalia alata  Irvingia malayana  Quercus kerrii	C = 0.022751 D2.209 H 0.458	11.2-58.2
9	Xylia xylocarpa  Dalbergia oliveri	C = 0.017539 D2.276 H 0.547	13.2-66.8
10	Quercus SP. Terminalia corticosa	C = 0.005957 D2.206 H 0.819	13.2-66.5

 

5.3 Tree Bole Carbon Equations in the Dry Evergreen Forest

Ten tree bole carbon equation derived from the Dry Evergreen Forest were constructed based on wood density, ranged between 387-1,257 kg/m³ (Table 3). A general equation which was used for all tree species in the dry evergreen forest was also constructed as followed (Eq. 3):

 

C = 0.011803 D^2.1844 H ^0.617…………………………..(Eq. 3)

 

                      Where;   C = Carbon storage in stem bole, kg/tree

                                    D = Diameter at breathn height of tree, cm

                                    H = Total height of tree, m

                                    

Table 3 The carbon equations classified by wood density of tree species in Dry Evergreen Forest

 

No.	Sample Species	Carbon Equations	DBH Range (cm)
1	Parkia leiophylla Tetrameles nudiflora  Duabanga grandiflora .	C = 0.049317 D1.997 H 0.357	18-147
2	Adenanthera pavonina  Cleidion spiciflorum  Croton roxburghii Podocarpus neriifolius Bischofia javanica	C = 0.019498 D2.300 H 0.300	12.5-42
3	Lithocarpus annamensis  Castanopsis acuminatissima Harpullia arborea Careya sphaerica	C = 0.012134 D2.056 H 0.668	12.0-3830
4	Artocarpus lakoocha  Terminalia nigrovenulosa  Dipterocarpus costatus  Eugenia aequea  Lagerstroemia tomentosa	C = 0.001549 D2.608 H 0.854	11.10-47.30
5	Dillenia obovata Cratoxylum formosum  Hopea odorata  Schima wallichii	C = 0.003192 D2.374 H 0.876	9.7-26.2
6	Anogeissus acuminata	C = 0.015560 D2.109 H 0.625	18.6-71.7
7	Pterocarpus macrocarpus	C = 0.014093 D2.068 H 0.723	11.5-61.5
8	Terminalia alata	C = 0.002624 D2.263 H 1.086	12.8-52.7
9	Xylia xylocarpa  Dalbergia cultrata Dalbergia oliveri Terminalia nigrovenulosa	C = 0.049317 D1.997 H 0.357	13.2-66.8
10	Quercus SP. Quercus lamellosa  Quercus kerrii  Terminalia corticosa	C = 0.006353 D2.482 H 0.609	10.9-43.7

 

 

5.4 Tree Bole Carbon Equation of Sector Mae Huad,Ngao Demonstration Forest

In oder to select the optimal tree bole carbon equation in all species of the Mae Huad Sector of the Ngao Demonstration Forest, Lampang Province, the general tree bole carbon equation (Eq. 4) is as follows:

 

C = 0.012348 D^2.1676 H ^0.6539…………………………..(Eq. 4)

           

                        Where;             C = Carbon storage in stem bole, kg/tree

                                                D = Diameter at breathn height of tree, cm

                                                H = Total height of tree, m

6. Estimation of carbon sequestration in each sampling point

Based on the equation to calculate carbon sequestration in each sampling point, the carbon storage per hectare were shown below (Table 1);

 

Table 1 Tree Bole Carbon Storage per Prism Plot

 

NO. of sampling point	Basal area per Hectare (m2/ha)	Tree per hectare (trees/ha)	Carbon per hectare (kg/ha)	Forest Type
1	33.69	407.86	66,448.65	MDF
2	21.88	256.65	8,638.54	DEF
3	27.56	1,454.44	57,241.25	MDF
4	–	–	–	AF
5	7.81	71.17	2,668.06	DEF
6	42.88	599.07	102,484.69	DDF
7	33.69	308.70	70,177.85	DDF
8	–	–		AF
9	30.63	1,659.10	47,420.85	MDF
10	–			AF
11	10.94	106.32	3,162.46	DEF
12	24.50	6,924.34	22,792.13	MDF
13	21.44	152.54	49,848.24	MDF
14	18.38	928.98	43,410.05	MDF
15	24.50	235.79	40,045.16	MDF
16	–	–	–	AF
17	–	–	–	AF
18	18.38	1,991.65	17,828.84	MDF
19	10.94	16.75	8,369.52	DEF
20	21.44	200.52	46,081.53	MDF
21	33.69	1,101.04	55,170.85	DDF
22	9.19	402.62	15,667.43	MDF
23	–	–	–	AF
24	–	–	–	AF
25	21.44	851.84	25,602.76	MDF
26	12.25	68.06	41,406.07	MDF
27	39.81	1,023.93	86,522.84	DDF
28	6.13	398.65	23,943.11	MDF
29	18.38	372.36	7,270.90	MDF
30	24.50	532.67	47,739.29	MDF
31	30.63	635.10	72,240.35	DDF
32	15.31	137.26	30,295.51	MDF
33	15.31	78.23	28,855.45	MDF
34	24.50	368.03	45,647.85	MDF
35	21.44	273.27	49,614.20	MDF
36	15.63	14.67	14,055.01	DEF
37	36.75	544.14	83,916.86	DDF
38	12.50	16.13	7,212.61	DEF
39	21.44	138.66	16,949.97	MDF
40	9.19	545.13	43,570.82	MDF
41	23.44	188.40	8,463.44	DEF
42	24.50	46.49	84,776.34	MDF
43	27.56	1,048.98	49,737.65	MDF
44	–	–	–	AF
45	36.75	1,263.64	74,931.79	DDF
46	–	–	–	AF
47	18.38	552.07	30,514.44	MDF
48	30.63	280.33	57,412.93	MDF
49	27.56	1,223.44	45,449.43	MDF
50	33.69	440.49	72,233.96	DDF
51	21.44	580.68	43,186.27	MDF
52	15.31	76.06	37,770.55	MDF
53	–	–	–	AF
54	17.19	28.07	15,759.18	DEF
SUM	993.13	28,544.33	1,802,535.68
AVERAGE	18.39	538.57	34,664.15

 

7. The carbon stock map generation 

7.1. Acquire and classify remote sensing data

The Normalized Difference Vegetation Index (NDVI) generated by utilized Landsat-8 images were acquired during the late of 2015, 2016 and 2017, to select the best output for further application i.e., linkage between NDVI and sample carbon plots, on a systematic grid (Figure 1).

Figure 1 A set of NDVI maps, with 30 meters resolution, based on Landsat – 8 data during 2015 to 2017.

7.2 The carbon plots transformation

. The total of 54 carbon plots (Table 1) were applied for carbon stock mapping. Adding carbon plots which known coordinate can be done via Excel table through GIS software.

Table 1 Illustrate part of the table prepared for carbon linkage

 

7.3 Carbon plot extraction and regression analysis

NDVI of different date were paired to a set of carbon plots of the same location to see the fitness between the carbon plots and NDVI Values. In Figure 2 revealed different scattering plots between carbon plots and NDVI raster value. The best fit was derived from January 16, 2017 and produced highest coefficient of determination (Figure 3). The regression model in the form of linear equation is;

 

                        C         =         154.68 NDVI -72.991             

when    R²        =         0.2297

                        R²        =         Coefficient of Determination

                        C         =         Carbon Storage in Tons

                        NDVI  =         Normalized Difference Vegetation Index Raster Value

Figure 2 Scattering plots of six carbon regression equations of 2017, reveal relationship between carbon and NDVI raster value extracted from the NDVI and stand carbon plots

Figure 3 A linear regression model used for carbon simulation in the study area

7.4 Collect and compile ground mapping data

The carbon storage gain from the 2^nd carbon coverage estimation details were illustrated in Table 2 and Figure 4 & 5. Reclassify of carbon generation using 2^nd inventory data revealed the different classes value when compared to those first inventory. Then, effect to carbon coverage area and carbon amount fall in each class. The grand total of carbon contained in the map is 1,714,496.879033 Tons. However, the carbon grand total is quite close to the first inventory, when the first map is estimated at  1,638,728.9157129 Tons of carbon.

Table 2 Average carbon storage (Tons) in 6 carbon classes after 2^nd inventory

Figure 4 Average carbon storage (Tons) in 6 carbon classes and the total after 2^nd inventory.

Figure 5  Carbon stock map of Mae Huad Sector, Ngao Demonstration Forest, Ngao District, Lampang Province, generated from the 2^nd inventory.

8. Quality assessment of the carbon map

The statistic analysis provide a satisfactory result, when the mean extract (read) from the carbon map is 34.339 while the mean indicate by ground inventory is 36.976, then, the  ratio between the mean map carbon estimates and mean ground estimates is equal 1.0768 or the map shall provide 92.87 % of the ground estimate. The others descriptive statistics is depicted in Table 3.

 

Table 3 Others descriptive statistics between carbon map and ground estimated

 

Pre (1st) Inventory Carbon (Tons)		Post (2nd) Inventory Carbon (Tons)
Column1		Column2
Mean	34.339	Mean	36.976
Standard Error	2.5584	Standard Error	2.9507
Median	37.641	Median	43.087
Mode	41.639	Mode	0
Standard Deviation	17.162	Standard Deviation	19.794
Sample Variance	294.53	Sample Variance	391.79
Kurtosis	9.405	Kurtosis	-0.281
Skewness	-2.859	Skewness	-0.69
Range	85.555	Range	72.928
Minimum	-34.95	Minimum	0
Maximum	50.608	Maximum	72.928
Sum	1545.3	Sum	1663.9
Count	45	Count	45
Confidence Level(95.0%)	5.156	Confidence Level(95.0%)	5.9467

 

The new carbon sample plots were compared to the carbon displayed on the carbon stock map. The statistics applied for the accuracy testing are descriptive statistics, F-test, Student’s t-test and the ratio between the mean ground estimates and the mean map carbon estimates. When testing for accuracy, the variance between two data sources was unequal, then statistical tests to determine the statistically significant by student’s t-test also assumed to unequal variance. The t-value calculated is less than the t-value in the table at alpha is 0.01 or it is considered no differences between the two estimates, at very high significant (Table 4).

Table 4 Statistic t-Test result of carbon stock map and ground estimated

t-Test: Two-Sample Assuming Unequal Variances
	Variable 1	Variable 2
Mean	36.97555809	34.33928299
Variance	391.7874863	294.5347995
Observations	45	45
Hypothesized Mean Difference	0
df	86
t Stat	0.675045325
P(T<=t) one-tail	0.25072893
t Critical one-tail	2.370493226
P(T<=t) two-tail	0.50145786
t Critical two-tail	2.634212309