We apologize for possible errors in the machine translation of the site

Flood Risk Mitigation (InVEST UFRM model)

PDF

GIS modeling and analysis, presentation of results – Eduard Kazakov (NextGIS OÜ, Estonia);
Analysis and presentation of results – Elena Bukvareva (BCC-Armenia);
Assessment of the alignment of results with real watersheds in Armenia – Vardan Asatryan (NAS RA) 
Assessment of the relevance of modeling results for Armenia – external expert Alexander Arakelyan (NAS RA) 

The InVEST model Urban Flood Risk Mitigation (UFRM) calculates two main indicators:
– the Runoff retention, i.e. the amount of runoff retained by soil and vegetation compared to the storm rain volume in m3 (1 m3 of runoff is equal to 10 mm per one 10x10m pixel) and as proportion of retained precipitation);
– the Runoff (Q), mm, which is a potentially hazardous factor that can cause flooding.

This section presents preliminary results of testing models for assessing and mapping ecosystem services. In the future, if a decision is made to use these models, they should be calibrated using hydrological measurements made in Armenia.
The methodology and results will be described in detail in a forthcoming publication.

ES provided by terrestrial ecosystems

We tested the model for two scenarios—average and extreme spring rainfall. The highest precipitation in Armenia occurs in May and June. While precipitation levels vary significantly across different climatic zones, for the initial model testing, we considered it reasonable to use countrywide average values. During these months, an average rainfall event delivers 12 mm of precipitation. For the extreme rainfall scenario, we assumed approximately half of the monthly precipitation in either of these months, which is 50 mm (Table 31A4-1).

Table 31A4-1. Precipitation and the number of days with rain in selected cities (http://armenia.pogoda360.ru/¹)

Climate zones	Cities	May			June
		Days with rain	Precipi-tation, mm	Average rain, mm	Days with rain	Precipi-tation, mm	Average rain, mm	Catastrophic rain, mm (50% of monthly precipitation)
Moderate cool	Sevan	12	140	12	13	157	12	79
	Hrazdan	10	113	11	10	120	12	60
	Stepanavan	11	141	13	10	130	13	65
	Vanadzor	13	177	14	13	189	15	95
	Average	12	143	12	12	149	13	75
Moderate relatively humid	Idjevan	10	127	13	8	97	12	64
	Dilijan	11	133	12	12	133	11	67
	Alaverdi	10	134	13	8	100	13	67
	Goris	9	103	11	5	63	13	52
	Average	10	124	12	8	98	12	62
Arid	Armavir	7	33	5	7	28	4	17
	Ararat	2	39	20	1	20	20	20
	Meghri	6	81	14	3	44	15	41
	Average	5	51	10	4	31	8	26
Moderate with dry summer	Gyumri	6	78	13	5	71	14	39
	Gavar	13	147	11	13	166	13	74
	Vardenis	9	109	12	9	99	11	55
	Sisian	8	112	14	7	84	12	56
	Average	9	112	12	9	105	12	56
Average		9	110	12	8	99	12	48

ES provided by terrestrial ecosystems: the average spring rainfall scenario (12 mm)

ES maps (Figure 31A4-1) show that precipitation is almost entirely retained by vegetation and soil. Quick runoff across most of Armenia is less than 1 mm, slightly exceeding this value in some valleys. Under the bare ground scenario (all natural vegetation is replaced with bare soil), runoff retention (RT) reduces very slightly. Quick runoff (Q) increases slightly in absolute terms, but the relative changes in some watersheds are noticeable.
Figure 31A4-1. Maps of ES indicators under the average spring rainfall scenario (12 mm). For detailed maps see the section “Ecosystem Services – Urban Flood Risk Mitigation – Average rainfall (12 mm) in the progect WebGIS

The ES provided by natural terrestrial ecosystems estimated as the difference in indicators between the ES on current land cover 2023 and on bare ground scenario. The influence of ecosystems on ES indicators is minor, amounting to a decrease in quick runoff by 0.01–0.08 mm and an increase in runoff retention by 1–8 liters per pixel; for the Hrazdan watershed a very small but opposite effect is observed. In relative terms, the effect on runoff retention is extremely small—everywhere under 1% of the 2023 value—and there is a wide spread in the share of quick runoff changes, ranging from +55% to -3%. (Table 31A4-2; Figures 31A4-2 and 31A4-3).

Figure 31A4-2. Ecosystem effect on quick runoff and runoff retention across watersheds under the average spring rainfall scenario (12 mm).Figure 31A4-3. Ecosystem effect on quick runoff and runoff retention reltive to ES on current land cover (2023), %

Table 31A4-2. ES indicators across watersheds under the average rainfall scenario (12 mm)

	Indicator	Aghstev	Akhuryan	Arpa	Debed	Hrazdan	Metsamor	Vorotan
Current land cover, ESRI 2023	Quick flow, mm, Q2023	0.11	0.40	0.16	0.10	0.24	0.27	0.07
	Runoff retention, m3/pix, RT2023	1.19	1.16	1.18	1.19	1.18	1.17	1.19
	Total runoff retention, mln of m3, RT2023Tot	37.76	32.07	52.19	46.71	70.43	42.78	53.42
Bare ground scenario	Quick flow, mm, Qbare	0.19	0.41	0.16	0.16	0.26	0.26	0.13
	Runoff retention, m3/pix, RTbg	1.18	1.16	1.18	1.18	1.17	1.17	1.19
	Total runoff retention, mln of m3, RTbgTot	37.51	32.04	52.16	46.49	70.30	42.81	53.16
Effect of terrestrial eco-systems	Reduction of quick runoff by ecosys-tems, mm Qeco = Q2023-Qbg	-0.06	-0.08	-0.01	-0.06	0.01	-0.02	-0.01
	Share of Q reduced by ecosystems, % Qeco*100/ Q2023	-54.88	-19.81	-3.94	-54.91	3.05	-8.13	-13.14
	Runoff retention, provided by ecosystems, m3/pix RTeco = RT2023-RTbg	0.01	0.01	0.00	0.01	0.00	0.00	0.00
	Share of RT provided by ecosystems, % RTeco*100/ RT2023	0.50	0.69	0.05	0.47	-0.06	0.18	0.08
	Total runoff retention, provided by ecosystems, mln of m3 RTecoTot = RT2023Tot -RTbgTot	0.26	0.25	0.03	0.22	-0.03	0.13	0.03
	Share of total RT, provided by ecosystems RTecoTot*100/RT2023Tot	0.70	0.79	0.05	0.47	-0.04	0.30	0.05

ES provided by terrestrial ecosystems: the extreme spring rainfall scenario (50 mm)

Precipitation is fully retained only in a small part of the territory (the darkest areas on the map of runoff retention). As a result, quick runoff exceeds 10 mm across most of the territory and exceeds 20 mm in a significant portion. If all natural vegetation is replaced with bare ground, runoff retention decreases significantly, and quick runoff also increases noticeably. Unlike the average-rain scenario, under an extreme-rain event the ecosystems’ influence on the ES indicators is substantial: they reduce quick runoff by an average of 4 mm (−32% relative to the value in 2023) and increase runoff retention by 0.4 m³/pix (+11% relative to the value in 2023). Totally, ecosystems increase runoff retention by 118 millions of m³
Figure 31A4-4. Maps of ES indicators under the extreme spring rainfall scenario (50 mm). For detailed maps see the section “Ecosystem Services – Urban Flood Risk Mitigation – Extreme rainfall (50 mm) in the progect WebGIS

Ecosystems increase runoff retention across all watersheds by 0.3–0.5 m³ and reduce quick runoff by 2.9–5.3 mm (Fig. 31A4-5; Table 31A4-3). In relative terms, compared to 2023 values, the ecosystem effect is most pronounced in the Arpa and Vorotan watersheds, where runoff retention increased by 13% and quick runoff decreased by 43–49% (Fig. 31A4-6; Table 31A4-3).

Figure 31A4-5. ES indicators under the extreme spring rainfall scenario (50 mm) across watersheds

Figure 31A4-6. Ecosystem effect: percentage change in runoff retention (R) and quick runoff (P) relative to 2023, by watershed

Table 31A4-3. ES indicators across watersheds under the extreme rainfall scenario (50 mm)

	Indicator	Aghstev	Akhu–ryan	Arpa	Debed	Hraz–dan	Metsa–mor	Voro–tan
Current land cover, ESRI 2023	Quick flow, mm, Q2023	13.3	16.6	10.8	13.3	13.4	14.3	11.7
	Runoff retention, m3/pix, RT2023	3.7	3.3	3.9	3.7	3.7	3.6	3.8
	Total runoff retention, mln of m3, RT2023Tot	116	92	173	144	219	130	172
Bare ground scenario	Quick flow, mm, Qbare	17.7	19.5	16.0	17.6	17.1	17.4	16.7
	Runoff retention, m3/pix, RTbg	3.2	3.1	3.4	3.2	3.3	3.3	3.3
	Total runoff retention, mln of m3, RTbgTot	103	84	150	127	197	119	149
Effect of terrestrial eco-systems	Reduction of quick runoff by ecosystems, mm Qeco = Q2023 -Qbg	-4.4	-2.9	-5.3	-4.3	-3.7	-3.1	-5.1
	Share of Q reduced by ecosystems, % Qeco*100/ Q2023	-32.8	-17.4	-49.0	-32.0	-27.5	-21.8	-43.3
	Runoff retention provided by ecosystems, m3/pix RTeco = RT2023 -RTbg	0.4	0.3	0.5	0.4	0.4	0.3	0.5
	Share of RT provided by ecosystems, % RTeco*100/ RT2023	11.9	8.6	13.4	11.6	10.1	8.7	13.2
	Total runoff retention, provided by ecosystems, mln of m3 RTecoTot = RT2023Tot -RTbgTot	14	8	23	17	22	11	23
	Share of total RT, provided by ecosystems RTecoTot*100/RT2023Tot	11.9	8.6	13.4	11.6	10.1	8.7	13.2

Changes in ES

Land-cover changes from 2017 to 2023, as captured in ESRI data, resulted in negative changes across all watersheds except Arpa. The most pronounced negative changes are modeled for the Akhuryan watershed, where runoff retention decreased by 1.5% and quick runoff increased by 3.8%. In the other watersheds (except Arpa), runoff retention decreased by 0.1–0.7%, while quick runoff increased by 0.3–1.5% (Figure31A4-7; Table 31A4-4). Changes in ES at the marz level mirror those at the watershed level. The changes are negative everywhere except in Vayots Dzor marz. The most pronounced negative changes are modeled for Shirak marz, which lies within the Akhuryan watershed (Figure 31A4-8; Table 31A4-4).Figure 31A4-7. Changes in ES under the extreme rainfall scenaio (50 mm) from 2017 to 2023 across watersheds

Figure 31A4-8. Changes in ES under the extreme rainfall scenaio (50 mm) from 2017 to 2023 across marzes

Table 31A4-4. Changes in ES under the extreme rainfall scenaio (50 mm) from 2017 to 2023

EAA		Changes in absolute terms		Changes relative to the values in 2017, %
		Quick runoff, Q, mm	Runoff retention, RT, m3/pix	Quick runoff, Q	Runoff retention, RT
Watersheds	Aghstev	0.037	-0.001	0.315	-0.096
	Akhuryan	0.545	-0.011	3.822	-1.526
	Arpa	-0.012	0.000	-0.091	0.034
	Debed	0.168	-0.003	1.262	-0.460
	Hrazdan	0.147	-0.003	1.362	-0.373
	Metsamor	0.243	-0.005	1.461	-0.727
	Vorotan	0.083	-0.002	0.619	-0.225
Marzes	Aragatsotn	0.082	-0.008	0.706	-0.213
	Ararat	-0.016	0.002	-0.157	0.041
	Armavir	0.042	-0.004	0.357	-0.110
	Gegharkunik	0.182	-0.018	1.180	-0.527
	Kotayk	0.125	-0.012	1.089	-0.324
	Lori	0.662	-0.066	2.901	-2.438
	Shirak	0.234	-0.023	1.720	-0.643
	Syunik	0.077	-0.008	0.532	-0.217
	Tavush	0.590	-0.059	3.722	-1.728
	Vayots Dzor	0.012	-0.001	0.084	-0.033

Using this ES as a case study, we tested the feasibility of assessing ES loss resulting from the historical conversion of natural grasslands by humans. The loss was assessed as the difference between the ES indicator values for the 2023 land cover and for a fully natural land-cover scenario in which all croplands and built-up areas are replaced by grasslands. ES loss is greatest—both in absolute and relative terms—in the Akhuryan watershed (a 5% decrease in runoff retention and a 10% increase in quick runoff), and smallest in the Arpa watershed (−0.7% and +2.7%, respectively) (Figure 31A4-9). Nonetheless, the results suggest that the ES has been mostly retained.

Figure 31A4-9. ES loss resulting from the historical conversion of natural grasslands by

As expected, the most significant loss of ES occurred in areas that are currently built-up where quick runoff increased the most—by 49%. For croplands, the ES loss is less substantial (Figure 31A4-10).

Figure 31A4-10. ES loss loss in built-up areas and in croplands